feasibility studies concrete in hydraulic structures: … · 2017-03-28 · tenance and restoration...
TRANSCRIPT
REPAIR, EVALUATION, MAINTENANCE, AND REHABILITATION RESEARCH PROGRAM
TECHNICAL REPORT REMR-CS-6
IN SITU REPAIR OF Dt:TERIORATED CONCRETE IN HYDRAULIC STRUCTURES:
FEASIBILITY STUDIES
by
R. P. Webster, L. E. Kukacka
Process Sciences Division Brookhaven National Laboratory
Upton, New York 11973
May 1987
Final Report
Approved For Public Release; Distribution Unlimited
Prepared tor DEPARTMENT OF THE ARMY US Army Corps of Engineers Washington, DC 20314-1000
Under Support Agreement No. WESSC-85-02 (Work Unit 32308)
Monitored by Structures Laboratory US Army Engineer Waterways Experiment Station PO Box 631, Vicksburg, Mississippi 39180-0631
The following two letters used as part of the number designating technical reports of research published under the Repair, Evaluation. Maintenance, and Rehabilitation (REMR) Research Program identify the problem area under which the report was prepared:
cs GT
HY
co
Problem Area
Concrete and Steel Structures
Geotechnical
Hydraulics
Coastal
EM
El
OM
Problem Area
Electrical and Mechanical
Environmental Impacts
Operations Management
For example, Technical Report REMR-CS-1 is the first report published under the Concrete and Steel Structures problem area.
COVER PHOTOS:
Destroy this report when no longer needed. Do not return it to the originator.
The findings in this report are not to be construed as an official Department of the Army position unless so designated
by other authorized documents.
The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of
such commercial products.
TOP- Deteriorated concrete gate pier, Montgomery Dam
BOTTOM- Pressure injection of concrete cracks, Bloomington Dam Intake
Unclassified SECURITY CLASSIFICATiON OF tA1s PAGE
REPORT DOCUMENTATION PAGE ,.$: ••
1a. REPORT SECURITY CLASSIFICATION 1 b. RESTRICTIV~ MARKINGS
Unclassified 2a. SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION I AVAILABILITY OF REPORT
2b. DECLASSIFICATION I DOWNGRADING SCHEDULE Approved for public release; distribution unlimited.
4. PERFORMING ORGANIZATION REPORT NUMBER(S) 5. MONITORING ORGANIZATION REPORT NUMBER(S)
BNL 37881-R Technical Report REMR-CS-6
6a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
Brookhaven National Laboratory (If applicablt) US Army Engineer Waterways Experiment
Station 6c. ADDRESS (Oty, Star., and ZIP Coda) 7b. ADDRESS (City, Statt, and i!IP COO.)
PO Box 631 Upton, NY 11973 Vicksburg, MS 39180-0631
Ba. NAME OF FUNDING/SPONSORING &b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER ORGANIZATION (If applicabM)
US Army Coros of Engineers Support Agreement No. WESSC-85-02 Be. ADDRESS (City, Statt, and ZIP Coda) 10. SOURCE OF FUNDING NUMBERS
PROGRAM PROJECT TASK WORK UNIT
Washington, DC 20314-1000 ELEMENT NO. NO. NO. ACCESSION NO.
32308 11. TITLE (lncluch Security Clallification)
In Situ Repair of Deteriorated Concrete in Hydraulic Structures: Feasibility Study
12. PERSONAL AUTHOR(S)
Webster R. P. Kukacka. L. E. 13a. TYPE OF REPORT r13b. TIME COVERED 14. DATE OF REPO~T (Y•ar, Month, Day) rs. PAGE COUNT
Final reoort FROM TO May 1987 80 16. SUPPLEMENTARY NOTATION This is a report of the Concrete and Stee;L §tructures problem area or
the Regair Evaluation Maintenance, and Rehabilitation (REMR) Research Program. Re~ort is ~zf~ta le from National Technical Information Service, 5285 Port Royal Road, SpringfLeld, VA
17. COSA Tl CODES 1 a. SUBJECT TERMS (Continu• on r•v•rs• If n•c•ssary and id•ntify by bloclc numbar)
FIELD GROUP SUB·GROUP Concrete In situ repair Pressure injection Cracking Polymer impregnation Spalling Hydraulic structures Post reinforcement
'9. ABSTRACT (Continu• on ravtrw if nactaaty and idantify by block numbar)
Presented are the results of a program conducted to identify the repair methods and materials currently being used to repair and rehabilitate concrete structures deteriorating as a result of cracking and spalling. These repair methods and materials were evaluated for their applicability to the in situ repair of concrete hydraulic structures. From this evaluation, the following three repair techniques were identified as best suited for in situ repair procedures: pressure injection, polymer impregnation, and addition of reinforcement. Case histories illustrating the application of each technique are presented. Recommenda-tions are made for work to be performed to make these systems more applicable to the repair of hydraulic structures.
20. DISTRIBUTION I AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATION
IXJ UNCLASSIFIED/UNLIMITED 0 SAME AS RPT. 0 OTIC USERS Unclassified 22a. NAME OF RESPONSIBLE INDIVIDUAL
DO FORM 1473, B4 MAR
22b. TELEPHONE (lncluch Arn COO.) 22c. OFFICE SYMBOL
B3 APR ad1t1on may ba used until txhausttd. All other aditions .,,, obsolete.
SECURITY CLASSIFICATION OF THIS PAGE
Unclassified
PREFACE
The study reported herein was authorized by Headquarters, U.S. Army
Corps of Engineers (HQUSACE), under Civil Works Research Work Unit 32308,
"In Situ Repair of Deteriorated Concrete," for which Mr. James E. McDonald
is principal investigator. This work unit is part of the Concrete and
Steel Structures Problem area of the Repair, Evaluation, Maintenance, and
Rehabilitation (REMR) Research Program. The Overview Committee of HQUSACE
for the P~MR Research Program consists of Mr. James E. Crews, Mr. Bruce
L. McCartney, and Dr. Tony C. Liu. Technical Monitor for this study was
Dr. Liu.
This study was monitored by the U.S. Army Engineer Waterways Experi
ment Station (WES) and conducted by the Brookhaven National Laboratory
(BNL) under the auspices of the U.S. Department of Energy under Sup-
port Agreement No. WESSC-85-02. This report was prepared by Messrs.
R. P. Webster and L. E. Kukacka, Process Sciences Division, BNL. The
study was performed under the general supervision of Mr. Bryant Mather,
Chief, Structures Laboratory (SL), and Mr. John M. Scanlon, Chief, Con
crete Technology Division (CTD), SL, and under the direct supervision of
Mr. McDonald, CTD. Program Manager for REMR is Mr. William F. McCleese,
CTD.
COL Dwayne G. Lee, CE, is Commander and Director of WES. Dr. Robert W.
Whalin is Technical Director.
1
CONTENTS
PREFACE • • • • • •
Page
1
CONVERSION FACTORS, NON-SI TO SI (METRIC) UNITS OF MEASUREMENT 4
PART I: INTRODUCTION • • • . • • • • • • • . 5
Background • • • • • 5 Program Objectives • • • • • • 6
PART II: EVALUATION OF REPAIR METHODS 7
Background . . . . . . . . . . . . • • • . 7 Repair and Rehabilitation Needs • . • • • • • • • 7 Methods and Materials for the Repair of Cracks in Concrete • 11
Pressure Injection Routing and Sealing • • • Stitching • • • • • • • Additional Reinforcement Drilling and Grouting Flexible Sealing
. • • • • . . . • 11 . . . . . . . 12
. . . . . . . 12 . . . . . . . . 14
. . . . . . . . . . . . . 14 . . . . . . . . . . . . . . 16
Grouting . . . . . . . . . 16 Drypack Mortar • • • • Crack Arrest
. . . . . . . 17 . . . . . . . . . . . 17
Polymer Impregnation • • • • Overlays and Surface Treatments • Autogenous Healing • • • •
. . . . . . 18
Methods and Materials for Repair of Spalled Concrete •
Repair Methods
Coatings • • • • • • • • • Concrete Replacement • • • • • • •
• • • 18 • 19
• • • • 19
• • 19
• 19 • • • • 20
Grinding • • • . • • • • • Jacketing • • • • • • . • • • •
. . . . . . . . 20 . . . . . 20
Shotcreting • • • • • • • • . . . . . . 20 Preplaced-Aggregate Concrete Thin-Bonded and Unbonded Overlays
. . . . . . 20 . . . . . . . . . 21
Repair Materials • 21
Bituminous Coatings Concrete, Mortar, and Grout Epoxies • • • • . • • • •
. . . . . . . . . . 21 • • 22
Expanding Mortars, Grouts, Linseed Oil • • • • • • • Latex-Modified Concrete
• • . 22 and Concretes • • • • • • 22
Polymer-Concrete Materials
Selection of Recommended In Situ Repair Techniques
Pressure Injection • • • • Polymer Impregnation • • • • • • • • • • Addition of Reinforcement •••.• Thin Reinforced Overlays and Shotcrete •
2
• • • 23 • • • 23
• • • • 23
• • • • 24
• • • • 24 • • • • 25
• 27 • • • 27
PART III: CASE HISTORIES
Pressure Injection •
Daniel Johnson Dam Pacoima Dam • • • •
CONTENTS
Twin Lakes Dam • • . • . • • • Lock and Dam 20, Pier 39, Canton, Missouri Equipment, Labor, and Economics •••••••
Polymer Impregnation • •
Page
• 29
• • • • 29
• • 29 • 30
. • 31 • • • • • • 32
. 40
• • • 41
Cass County Jail, Fargo, North Dakota • • •••• 41 Greenport Bridge, Greenport, NY • • • • • 43 Dworshak Dam, Orofino, Idaho • • • • • • . .••••• 46 Equipment and Economics • • • • • • • • • • • • 53
Addition of Reinforcement • • • • 54
PART IV:
REFERENCES
TABLES 1-3
John Day Navigation Lock, Oregon • • • • Markland Locks and Dam, Ohio River • • •. Apartment Building, Brussels, Belgium • Kansas Department of Transportation
SUMMARY AND RECOMMENDATIONS
3
• • 54 61 63
• • 64
• • • • • 69
••• 74
CONVERSION FACTORS, NON-S! TO SI (METRIC) UNITS OF MEASUREMENT
Non-S! units of measurement used in this report can be converted to SI
(metric) units as follows:
Multiply
cubic yards
Fahrenheit de2rees
feet
gallons (U.S. liquid)
gallons (U.S. liquid) per
square foot
inches
miles (U.S. statute)
pounds (force) per square inch
square feet
square miles
By
0.7645549
5/9
0.3048
3.785412
40.745836
25.4
1.609347
0.006894757
0.09290304
2.589998
To Obtain
cubic metres
Celsius degrees or kelvins*
metres
cubic metres
liters per square metre
millimetres
kilometres
megapascals
square metres
square kilometres
* To obtain Celsius (C) temperature readings from Fahrenheit (F) readings, use the following formula: C = (5/9)(F- 32). To obtain kelvin (K) readings, use: K = (5/9)(F - 32) + 273.15.
4
Background
IN SITU REPAIR OF DETERIORATED CONCRETE IN
HYDRAULIC STRUCTURES: FEASIBILITY STUDY
PART I: INTRODUCTION
Over the last 7S to 80 years, reinforced portland-cement concrete
has been used extensively in hydraulic structures, such as dams, spill
ways, lock chambers, and bridge support columns and piers. The Corps of
Engineers estimates that it now operates and maintains S36 dams and 260
lock chambers at S96 sites (Scanlon 1983). Of these, more than 40% are
over 30 years old and 29% were constructed before 1940. It is further
estimated that nearly half the 260 lock chambers will reach the end of
their SO-year design lives by the turn of the century. Over this same
period of time, waterborne traffic is expected to increase approximately
SO%. With the relatively limited new construction starts anticipated,
many of these structures will be kept in service well beyond their origi
nal design lives. Periodic inspections of these facilities reveal that a
large number of the older structures require significant maintenance, re
pair, and rehabilitation. In addition, the relatively new structures
must be maintained to ensure their continued service and operation.
In most cases, repairs to such structures entail removal of the de
teriorated concrete and replacement with new concrete. If deterioration
is not too severe, the time and costs of rehabilitation are generally
acceptable. If, however, the deterioration is extensive, it is often
necessary to remove substantial amounts of concrete and, in some cases,
to completely remove and replace the concrete section. Considerable
savings in time and cost for the rehabilitation of highly deteriorated
concrete structures would be realized if methods and materials could be
identified and developed to repair such structures without the extensive
removal of the deteriorated concrete. To this end, Brookhaven National
Laboratory (BNL) has carried out a program entitled "In Situ Repair of
Deteriorated Concrete in Hydraulic Structures."
5
Program Objectives
The specific objectives of the BNL program were to evaluate existing
methods and materials for use in the in situ repair of deteriorated con
crete hydraulic structures, as well as to identify new materials and de
velop new concepts.
6
PART II: EVALUATION OF REPAIR METHODS
Background
Information ~ertaining to methods and materials used to repair port
land-cement concrete hydraulic structures was obtained (a) from a compu
terized literature survey, and (b) from mail and telephone inquiries and
meetings with government agencies and private firms active in the main
tenance and restoration of concrete structures.
The objectives of the survey were twofold: (a) to identify the
forms of deterioration most prevalent in concrete hydraulic structures,
and (b) to identify existing methods and materials commonly used for the
repair and rehabilitation of concrete structures. Once this information
was collected,. it was evaluated to determine the applicability of the
various systems to the in situ repair of concrete hydraulic structures.
The results of this survey are presented below.
Repair and Rehabilitation Needs
In 1982, the Corps of Engineers initiated a program to develop
quantitative information on the condition of the concrete portions of the
Corps' civil works structures in an attempt to identify the most critical
long-term needs with respect to repair and rehabilitation. This program
was done, in part, by reviewing existing periodic inspection reports for
Corps locks and dams in order to obtain input data for a computerized
data base to be used to identify and evaluate trends in deterioration and
other problem areas in concrete hydraulic structures. The results of
this study (McDonald and Campbell, 1985) indicated that the most common
types of concrete deficiences were (a) cracking, (b) seepage, and (c)
spalling. These three general catagories of deficiencies accounted for
77% of the total of 10,096 deficiencies identified during the review of
available inspection reports. Concrete cracking was the deficiency most
often observed, accounting for 38% of the total. In situ repair proce
dures may not be readily applicable in the repair of seepage deficien
cies. However, deterioration due to cracking and spalling does seem to
be suited to the use of in situ repair procedures.
7
Cracking and spalling of concrete can have a number of causes
including temperature stresses, adverse chemical reactions, corrosion of
reinforcing steel, weathering (cycles of freezing c-md thawing), acciden
tal overloading, and differential mo',"'Tllent of the structure. Much of
the damage typically found in many of the structures operated by the
Corps upparently is due to deterioration resulting from the penetration
of water into exposed surfaces which are horizontal or which have
inadequate drainage followed by freezing and thawing (Figures 1-3).
Vertical surfaces made using non-air entrained concrete are equally sus
ceptible to this type of deterioration. The initial cracking in these
structures may actually have any of several different causes including
drying shrinkage, thermal stresses, alkali-aggregate reaction, and
stresses due to structural movement; however, once the cracking has
occurred, water ponding on the surfaces of the structure results in the
penetration of moisture into the concrete mass. Subsequent cycles of
freezing and thawing of the saturated concrete leads to progressive
deterioration of the concrete, aggravated, in some cases, by the fact
that non-air-entrained concrete was used in many of the older
structures.
Although initial deterioration due to cracking and spalling may
visually appear to be very severe, the concrete member generally will
remain intact until the deterioration becomes excessive, at which time
the concrete will begin to break apart. Once this stage is reached, it
is generally necessary to remove and replace the deteriorated concrete.
However, if the concrete can be repaired using in situ repair procedures
before it has deteriorated this far, it is possible to extend the
service life of the structure.
With this objective in mind, the emphasis of the program was
directed toward identifying repair methods and materials to be used for
the in situ repair of hydraulic concrete structures exhibiting
deterioration due to cracking and spalling.
8
a. General view of deterioration due to cracking.
b. Closeup view of top of gate pier.
Figure 1. Typical cracking deterioration in concrete gate piers.
9
Figure 2. Typical cracking in concrete arch.
Figure 3. Typical cracking in concrete wall.
10
Methods and Materials for the Repair of Cracks in Concrete
Despite the general belief that it is a very durable construction
material, concrete is very susceptible to deterioration due to cracking.
As a result, cracking is the maintenance and repair problem most fre
quently encountered in concrete structures. As previously mentioned,
causes for cracking include errors in design and detailing, temperature
stresses, chemical reactions, corrosion of embedded metals (reinforcing
steel), weathering (freezing and thawing), accidental overloadings and
differential movement of the structure. The large number of factors
that can cause concrete to crack indicates that no one repair procedure
is appropriate in all instances. To ensure success, the cause of the
cracking as well as the present condition of the crack must be taken
into account in selecting a repair procedure.
In 1984, Committee 224 of the American Concrete Institute released
a report detailing the causes, evaluation, and repair of cracks in con
crete structures, and identified 12 techniques most commonly used for
the repair of cracks in concrete (see Table 1). A discussion of each
technique is presented below.
1. Pressure Injection. This repair technique, which has been used
successfully to repair cracks varying in width from 0.002 to 0.003 in.
up to 0.25 in., consists of drilling holes at close intervals along the
length of the cracks, installing injection ports, sealing the surface of
the crack, and injecting an adhesive, usually epoxy, under pressure into
the crack. Injection progresses from hole to hole, normally beginning at
the lowest point, and continues until the entire length of the crack has
been filled. Injection has been used successfully to repair cracks in
bridges, buildings, dams, and other types of structures. However, unless
the crack is dormant, it will probably recur, usually elsewhere in the
structure. If the crack is active and it is desirable to seal the crack
while allowing continued movement at that location, it is necessary to
use a sealant or other material that allows the crack to function as a
joint. Injection can be used, within limits, against a hydraulic head,
provided the injection pressure is adjusted upward to counteract the
hydraulic head.
11
Injection techniques require an average degree of skill for satis
factory execution, and application of the technique is limited by ambient
temperature. If the part of the structure to be repaired is subject to
varying seasonal temperature changes, the width of the cracks can vary
considerably. If possible, repairs should be scheduled during the cooler
seasons, especially in the spring, when cracks are at their widest.
Cracks filled during cooler weather will be in compression, whereas
cracks filled during summer or early fall, will likely be in tension when
the structure cools during winter months.
2. Routing and Sealing. Routing and sealing is the simplest and
most common technique used to repair cracks that are dormant and have no
structural significance. It is not applicable for sealing cracks sub
jected to high hydrostatic pressures, except when the pressurized face
is being sealed, in which case some reduction in flow can be obtained,
nor is it applicable in instances where aesthetics are important since
the technique ''highlights" the crack being repaired.
The technique simply entails enlarging the crack along its exposed
face and filling and sealing it with a suitable joint sealant (Figure 4).
Choice of the sealant depends on how tight or permanent the seal must be.
Epoxy compounds are most commonly used, although hot-poured joint seal
ants have been used when watertightness is not required and appearance
of the repair is unimportant. Urethanes, which remain flexible through
large temperature variations, have been used successfully in cracks up
to 0.75 in. in width.
3. Stitching. This technique involves drilling holes on both sides
of the crack and grouting in U-shaped reinforcing bars (stitching dogs)
that span the crack (Figure 5). A non-shrink grout or an epoxy resin is
normally used to anchor the legs of the bars. Stitching is generally
used when it is necessary to reestablish tensile strength across major
cracks. Stitching does tend to stiffen the area being repaired, which
may accentuate the overall stiffness of the structure and cause cracks to
develop elsewhere. It is, therefore, generally necessary to strengthen
the adjacent sections using external reinforcement.
12
•<I • .. . . .. . ~ .'
• ·'if>. I
. ., .
A) ~IGINAL CRACK
~VE CUT WITH SAW OR
CHIPPING TOOL.
<1. . ... '
·. b .:·J,. .. ~ ..
B) ROUTING
. . ~· .. . : I 0 '
a
c) SEALING
Figure 4. Repair of crack by routing and sealing (ACI Committee 224, 1984).
NOTE VARIABLE LENGTH, LOCATION AND ORIENTATION OF DOGS SO THAT
TENSION ACROSS CRACK IS DISTRIBUTED IN THE CONCRETE RATHER
THAN COCENTRATEO ON A SINGLE PLANE,
HoLES ARE DRILLED INTO THE CONCRETE TO
RECEIVE THE STITCHING DOGS, WHICH
ARE CEMENTED INTO PLACE,
STITCHING DOGS
Figure 5. Repair of crack by stitching (ACI Committee 224, 1984).
13
Stitching will not close a crack but can be used to prevent it from
propagating further. If water is a problem, the crack should be made
water tight before it is stitched to protect the bars from corrosion.
4. Additional Reinforcement. Cracked reinforced concrete members
have been successfully repaired by adding reinforcement to the member,
both internally (post-reinforcement) and externally.
Post-reinforcement is a technique developed by the Kansas Depart
ment of Transportation to repair cracked structural bridge concrete. The
method consists of sealing the surface of the crack, drilling holes at
45° to the deck surface, and crossing the crack plane at approximately
90°, filling the hole and crack plane with epoxy pumped under low pres
sure, and placing a reinforcing bar into the drilled hole in a position
to span the crack (Figure 6). The epoxy bonds the bar to the walls of
the hole; it fills the crack plane, thereby rebonding the cracked con
crete surfaces in one monolithic form and reinforcing the section.
Externally applied post-tensioning is often used when a major por
tion of a member must be strengthened or when cracks need to be closed.
The technique uses prestressing tendons or bars to apply a compressive
force to the face of the member being repaired. The effects of the ten
sioning forces on the rest of the structure must be analyzed to prevent
cracks from developing in other parts of the structure.
5. Drilling and Grouting. This technique is useful for repairing
vertical cracks which are reasonably straight and are accessible at one
end of the crack such as in retaining walls. It consists of drilling a
hole, usually 2 to 3 in. in diameter, down the length of the crack and
grouting it in to form a key (Figure 7). The grout key prevents trans
verse movement of the concrete adjacent to the crack and will also reduce
heavy leakage through the crack. If watertightness is essential and
structural load transfer is not, the drilled hole is filled with a resil
ient material in place of the grout.
14
HOLES ARE DRILLED AT 45° TO DECK SURFACE, CROSSING
CRACK PLANE AT APPROXIMATELY q(JD, HoLE IS FILLED
WJTI1 EPOXY, UNDER LOW PRESSURE, AND REBAR IS
INSERTED INTO HOLE IN A POSITION TO SPAN CRACK,
Figure 6. Repair of crack by post reinforcement.
lli FORM KEY WITH PRECAST CONCRETE OR
MORTAR PLUGS SET IN BITI.MEN,
HoLE DRILLED IN STEM OF WALL, CENTERED
ON AND FOLLOWING DOWN CRACK, SIZE OF
HJLE DEPENDS ON WIDTH OF CRACK.
Figure 7. Repair of crack by drilling and plugging (ACI Committee 224, 1984).
15
>;_ ·_.o ..
-. d
0 -_ 'I ... ~- ..
0.
- d
.... ·o'
·.t1 b ..
0 .· ~ . n . \\
. J7 •. r_. . \> 0
4 . t:J
p'
~-·-----BACKUP MA.TERIAL
Figure 8. Repair of crack using a flexible seal (ACI Committee 224, 1984).
6. Flexible Sealing. Active cracks can be sealed by routing them out
and filling them with a sUitable flexible sealant. This technique is
applicable where appearance is unimportant and in areas where the cracks
are not subject to traffic or mechanical abuse. The repairs are made by
routing out a slot or groove along the length of the crack of sufficient
width and shape to accommodate the expected movement. The slot is then
filled with a backing material, covered with a bond breaker, and finally
sealed with the flexible joint sealant (Figure 8).
7. Grouting. Wide cracks, particularly in gravity dams and thick con
crete walls, can be repaired with portland-cement grout. Narrow cracks may
be repaired with chemical grouts consisting of solutions of chemicals that
combine to form a gel, a solid precipitate, or a foam. Cracks as narrow as
0.002 in. have been repaired using chemical grouts. The procedure, in
general, consists of cleaning the concrete along the length of the crack;
installing grout nipples at intervals, and sealing the crack between the
nipples; flushing the crack to clean it and to test the seal; and theq
pumping the grout. After the crack is filled, the pump pressure is main
tained for several minutes to ensure good penetration of the grout.
16
8. Drypack Mortar. This technique consists of the hand placement of
a low water content mortar followed by tamping or ramming of the mortar
into place to produce tight contact between the mortar and the existing
concrete. The repair exhibits very little shrinkage and the patch remains
tight, with good durability, strength, and watertightness. It is appro
priate for use in cavities that are deeper than they are wider. Since
formwork is not required, drypack is especially appropriate for use in
vertical members. It is not appropriate for the repair of extensive, wide
or shallow areas or for applications that require compaction of the mortar
behind obstacles, such as reinforcing bars. The technique is used for the
repair of dormant cracks but is not recommended for active cracks.
9. Crack Arrest. This technique is commonly used to prevent re
flective cracking into new concrete which is being placed over existing
concrete. The technique, in general, arrests the cracking by blocking
it and spreading the tensile stresses causing the cracking over a larger
area. One technique consists of placing semicircular sections of steel
pipe over the crack and then covering the pipe with concrete to anchor
it in place. After this concrete cap has set, the new concrete is placed
over the cap. The steel pipe is later filled with grout to complete the
structural continuity of the member (Figure 9).
(.ONCRETE PLACED CONCENTRICALLY OVER
PIPE AT LEAST 2 DAYS PRIOR TO PLACEMENT OF UPPER LIFT,
9. :; r ~ • '\7' • ;'--.• • • • /7 • • • • ••• Ot
~ . . P. ·.<r· .. 0 . . ''{ .· . . .. · ..
• !
I>'
~·
:-q- -~. "<:]' • :.. . • [>. - ' c>·_ . • • • t> :
. :· .. ~. ~·. t> 1> . . ..
. . . . .
~
. . . . . \>' .
.o
CRACK
Figure 9. Repair of crack using the crack arrest method (ACI Committee 224, 1984).
17
10. Polymer Impregnation. Surface impregnation techniques have
been used as a means for reducing chloride and moisture penetration into
concrete bridge decks by filling the pore structure in the concrete with
polymer. However, it has also been successfully used to restore the
structural integrity of highly deteriorated concrete bridge decks and
floor dabs as well as to improve the abrasion resistance of the concrete
in outlet tunnel walls of dams. The impregnation process consists of
four basic steps:
a) preparation of the surface to remove contaminants or films that would prevent or reduce monomer penetration,
b) drying the concrete to a depth sufficient to permit the desired monomer penetration,
c) impregnation of the concrete with liquid monomer to the desired depth,
d) polymerization of the monomer within the pores of the concrete.
Vacuum impregnation techniques are similar to surface impregnation
except that the liquid monomer is forced into the pore structure under
pressure, in this case a negative pressure, to produce greater penetra
tion of monomer into the concrete. It also provides a method for impreg
nating surfaces which are not compatible with the gravity-soak method of
impregnation such as overhead and vertical surfaces.
11. Overlays and Surface Treatments. Concrete containing fine dor
mant cracks can be repaired by applying a bonded overlay to the surface.
However, most cracks are subject to movement caused by variations in
loading, temperature, and moisture. Cracks subjected to such forces will
reflect through any bonded overlay, defeating the purpose of the overlay
insofar as repair of cracks is concerned. Bonded overlays are also sub
ject to cracking due to drying shrinkage and restraints at the inter
face. Unbonded overlays can be used to cover surfaces with moving
cracks. However, thin unbonded overlays which receive load may crack
due to a combination of thermal and load stresses.
18
Cracks may also be repaired by sealing the surface of the member
with a sealant such as an epoxy resin. However, concrete on grade in
freezing climates should never receive a surface treatment which will act
as a vapor barrier. This would allow moisture passing from the subgrade
or surrounding environment to condense under the barrier, leading to
critical saturation of the concrete and rapid deterioration by freezing
and thawing.
12. Autogenous Healing. Autogenous healing is a natural process of
crack repair that can occur in concrete in the presence of moisture and
the absence of tensile stresses. This healing process is dependent upon
the carbonation of calcium hydroxide in the cement paste by carbon diox
ide present in the surrounding air and water. Calcium carbonate and cal
cium hydroxide crystals form and grow within the cracks. The chemical
and mechanical bonding resulting between the crystals and the surfaces of
the paste and the aggregate restores some of the tensile strength of the
concrete across the cracked section, and the crack may become sealed.
Healing will not occur if the crack is subject to movement or if there is
a· positive flow of water through the crack.
Methods and Materials for Repairing Spalled Concrete
As with the repair of cracks in concrete, a number of methods and
materials have been used to repair concrete subjected to surface spalling
and scaling. This section covers the basic repair techniques and materi
als most commonly used (Concrete Construction, 1982). The techniques and
materials are presented in Tables 2 and 3 and are discussed below.
Repair Methods
1. Coatings. This technique is generally used when scaling or
spalling is limited to a very thin region at the surface of the concrete.
It consists of painting a film-forming plastic or liquid coating over the
surface of the concrete. Coatings can be applied by brushing, rolling,
or spraying. Some common applications for coatings are reduction of
ingress of water, protecting concrete from aggressive chemicals, and
providing a durable wearing surface under heavy traffic loads.
19
2. Concrete Replacement. Unsound concrete can be removed and
replaced with conventional portland-cement concrete or some other patch
ing material. Whether the existing concrete is completely or only
partially removed depends upon the extent and narnre of the deteriora
tion. This technique is one of th2 most commonly used and is appropriate
for applications where the cause of the deterioration is nonrepeating or
has been eliminated.
3. Grinding. This technique can be used when the deterioration is
limited to a thin region at the surface of the concrete. However, unless
modern heavy-duty equipment is used, it can be a slow, expensive, dusty
method of repair.
4. Jacketing. This technique entails fastening a material to the
existing concrete that is more resistant to the environment that is caus
ing the deterioration. The material can be metal, rubber, plastic, or
high-strength concrete; it can be secured to existing concrete by bolts,
nails, screws, adhesives, straps, or gravity. Common applications in
clude tanks, spillways, piers, and other concrete elements that are ex
posed to corrosive materials or rapidly flowing water.
5. Shotcreting. This technique entails shooting concrete or mortar
under pressure into the cavity or onto the surface of the concrete to be
repaired. It may involve pumping completely mixed material through the
hose or blowing the dry constituents through the hose and mixing them
with water at the nozzle. The latter method requires an experienced op
erator but offers the capability of customizing the shotcrete water con
tent and consistency to the needs of specific areas of the repair job.
Shotcrete is practical for large jobs, on either vertical or horizontal
surfaces, where the cavities are relatively shallow.
6. Preplaced-Aggregate Concrete. This technique which is also
called "prepacked" concrete, entails prepacking gap-graded aggregate into
the cavity to be repaired and inundating the cavity with water to satu
rate the aggregate. Mortar or grout is then pumped in from the bottom,
20
displacing the water. This technique is suitable for inaccessible appli
cations, such as submerged concrete or deteriorated concrete that is
being jacketed. Preplaced-aggregate concrete has been used to repair
piles, footings, piers, retaining walls, abutments, base plates, tunnels,
and dams. Preplaced-aggregate concrete provides low shrinkage and good
bonding qualities but can leave voids. Because of the specialized skills
and equipment required, this type of repair work is usually performed by
a firm that specializes in it.
7. Thin-Bonded and Unbonded Overlays. Thin overlays, i.e., over
lays of 2 in. or less, are often used to repair concrete surfaces that
are basically sound structurally but have deteriorated from cycles of
freezing and thawing, heavy traffic, or other exposures which the origi
nal concrete was unable to withstand. Overlays are also occasionally
used to relevel slabs or to reestablish grades.
Once a decision has been made to place an overlay, it must be
determined whether the overlay should be bonded or unbonded. If the
deterioration to be repaired is a surface phenomenon, such as spalling
or scaling, the overlay is usually bonded. In the case of cracking or
structural movement, however, it may be desirable not to bond the over
lay so that it will not reflect the distress in the base slab.
Repair Materials
1. Bituminous Coatings. Asphalt- or coal tar-based bituminous
coatings are used to reduce the tendency of water to enter concrete or to
protect it to some extent from weathering. They are low in cost, famil
iar to workmen, and effective for a time, if properly applied; and their
thickness, and therefore, their resistance to water passage and weather
ing, can be varied to suit the exposure. The disadvantages of bituminous
coatings include the need to replace or renew them periodically, the
messiness and the odors associated with their application, their tendency
to dry and crack, their sensitivity to ambient temperatures, and the
rapid destruction of the coatings when certain liquids, such as gasoline,
are spilled on them.
21
2. Concrete, Mortar, and Grout. Portland-cement concrete, mortar,
and grout have a number of advantages as repair materials, including
thermal properties similar to those of the existing concrete, similarity
in appearance, relatively low cost, easy availability, and familiarity.
Concrete is most often used for the complete replacement of sections
and deep cavities extending beyond the top layer of reinforcing steel.
Mortar is generally used for cavities of 1 1/2 in. or less, or for appli
cations that are too shallow for the coarse aggregate in concrete and
where the fluidity of grout is unnecessary or not wanted. Grout has the
advantages of being fluid and readily pumped even into areas that cannot
be seen. Grout can be used where clearances are minimal and where it is
necessary to reduce the likelihood of leaving major voids. Grouts, on
the other hand, have a higher water content and consequently undergo more
drying shrinkage than well-designed mortar or concrete.
3. Epoxies. Epoxies are acknowledged to be excellent repair
materials for selected applications. Epoxies are organic compounds that,
when mixed with a hardening agent, produce a tough, rapid-setting, and
rapid-hardening material that is chemically and physically stable, and
that is resistant to water penetration, crack formation, and many chemi
cals that attack concrete. They also have excellent adhesive properties,
and they may be modified. For example, they can be made rubbery for
elastomeric sealing or caulking applications, or fine hard aggregate can
be broadcast over the top of a coating to produce a skid-resistant sur
face. Epoxies have the following disadvantages: high cost, allergenic
effects on some workmen, and significant differences from concrete in
important physical properties such as coefficient of thermal expansion,
tensile strength, and flexural strength. Epoxies are most often used in
concrete repair work as an adhesive to bond new concrete to hardened
concrete, for patching, and for coatings.
4. Expanding Mortars, Grouts, and Concretes. These materials are
generally proprietary materials that counteract the problem of shrinkage
by incorporating ingredients which produce an expansion approximately
equal in magnitude to the expected drying shrinkage. Variations in the
effectiveness of these materials are such that performance data and
22
previous applications should be checked before use. Some materials have
been found to produce excellent long-lasting repairs while others have
exhibited lack of density and other deficiencies that could undermine
their effectiveness.
5. Linseed Oil. Linseed oil has been used to correct scaling which
is not severe enough to warrant using a coating or some other form of re
pair. Strictly speaking, linseed oil is not a repair material because it
is used to prevent or minimize additional scaling, not to repair existing
scaling. The linseed-oil solution is applied to the slab surface and
penetrates up to 1/8 in., providing a film of low permeability that re
sists infiltration by aggressive solutions while allowing water vapor to
escape.
6. Latex-Modified Concrete. Concrete and mortar that have been
modified through the addition of a latex admixture have been used suc
cesssfully to repair deteriorated floors and bridge decks. Latex-modi
fied concretes typically exhibit good bond strength to existing sound
concrete as well as high compressive and tensile strengths, and they are
more flexible than unmodified mixtures and are resistant to weathering,
alkalies, and dilute acids. Though more expensive than unmodified port
land-cement concrete, latex-modified concrete is low in cost compared
with many other types of repair materials, including those based on
epoxies. On the other hand, they are in some ways more difficult to
handle and finish than other materials since they set rapidly and form a
skin that tears, if troweled after the skin has formed.
7. Polymer Concrete Materials. Polymer concrete patching materials
have been used extensively in the repair of highway bridge decks and
pavements. Unlike normal portland-cement concrete, polymer ~oncrete con
tains no water and, generally, no portland cement. Instead, the blended
aggregate filler is bonded together in a polymer matrix. A number of
materials such as methyl methacrylate, unsaturated polyesters, vinyl
esters, polyurethane, and epoxy have been used as the polymer matrix.
Polymer concrete materials have a number of advantages over normal port
land-cement concrete including rapid curing characteristics, high early
strength, good bond strength, impermeability to water penetration, and
excellent resistance to freezing and thawing • Disadvantages include the
23
need to pay detailed attention to manufacturers' recommendations, careful
formulation necessary to avoid problems due to differences in the
physical properties of the polymer concrete and the adjacent portland
cement-concrete, the relatively hii',h cost and the f12Lmmability and
toxicity of some of the constituents.
Selection of Recommended ~ ~ Repair Techniques
From an evaluation of the repair techniques and materials summarized
above, five procedures--three crack repair techniques and two techniques
for repairing spalled concrete--have been identified as being the most
applicable for use as in situ repair procedures for concrete hydraulic
structures. The selected repair techniques include pressure injection,
pol~ner impregnation, and additional reinforcement. In conjunction with
these repair procedures, thin reinforced overlays and shotcrete can be
used to repair spalled concrete surfaces as well as to resurface struc
tures after the cracks have been repaired. Criteria considered in the
selection of the repair techniques identified above included the
following:
The repair technique should (a) effectively repair both dormant and
active cracks, (b) restore the structural properties of the cracked mem
ber, (c) provide a watertight repair of the cracks, (d) allow for the
repair of the member in the presence of moisture or hydraulic pressures,
(e) improve the durability of the concrete member, (f) prevent access of
corrosive agents into the concrete, and (g) leave no visible scars or
surface blemishes on the repaired member.
Since no single method of repair would satisfy each of the above
criteria, it was necessary to identify more than one method of repair.
Presented below is a discussion of each of the selected techniques.
1. Pressure Injection. Pressure injection, generally using epoxy
adhesives, has been used extensively for about 25 years to repair a vari
ety of concrete hydraulic structures including dams, spillway tunnels,
bridge piers, and water-retaining tanks. As a result, much of the ex
pertise, technology, materials, and equipment necessary for the success
ful application of the process already exists.
24
A principal advantage of the pressure injection technique is that it
seals cracks externally and internally as opposed to such techniques as
routing and sealing and flexible sealing, which only seal t.:he cracks ex
ternally. The advantage of repairing the cracks internally as opposed to
only sealing them externally is that complete sealing of the full depth
of the crack will eliminate areas in which moisture may collect, thereby
reducing the possibility of damage due to freezing and thawing. Internal
sealing also helps to restore the structural integrity of the member
being repaired. Cracks as fine as 0.002 in. have been repaired using
pressure injection techniques.
Another advantage of this technique is that with proper selection of
a water-compatible adhesive, cracks saturated with water can be repaired.
Pressure injection can be used, within limits, against a hydraulic head,
provided the injection pressure is adjusted upward to counteract the
pressure of the hydraulic head.
While pressure injection is generally very successful when done
properly, it does have some limitations. For example, the process may
leave scars on the surface of the member where the cracks have been in
jected, unless the surface of the member is refinished once the repairs
have been completed. The application of the process is generally limited
to members which have not yet begun to spall and to the repair of dormant
cracks. The process can perhaps be used for partial repair of active
cracks; however, continual movement of the structure will lead to new or
continued cracking.
Pressure injection, however, appears to be one of the most viable
methods available for the repair of severely cracked structures, such as
those shown in Figures 1 through 3.
2. Polymer Impregnation. Although research regarding the basic
material properties of polymer-impregnated concrete has been in progress
for about 17 years, practical in situ applications on existing structures
have taken place only in the last 8 to 10 years. Structures repaired and
25
rehabilitated with polymer impregnation include highway bridge decks,
structural floor slabs, outlet tunnel walls, and stilling basins. It
should be pointed out that most oi these applications have been experi
mental in nature, and routine applications of the process are yet to be
come commonplace.
Although the laboratory research and the field applications have
shown that polymer impregnation is an excellent method for rehabilitating
highly deteriorated concrete, the process has several limitations which
currently prevent its use from becoming routine. The limitations include
the following:
Because the process is relatively new and uses specialized materials
and equipment, a relatively high level of expertise and supervision is
required, and it cannot be performed by untrained or unsupervised person
nel. The monomer systems currently being used to impregnate concrete
require specialized safety procedures since they are often flammable and
toxic. In addition, the monomers are not water compatible. Therefore,
to ensure complete penetration of monomer into the pore structure of the
concrete, the concrete must be dried to remove the free moisture within
the pores. These limitations can make polymer impregnation a relatively
expensive method of repair, when compared to more conventional methods.
Despite these limitations and restrictions, polymer impregnation ap
pears to offer a very good method for rehabilitating or improving the
overall physical and mechanical characteristics of highly deteriorated,
low-quality, or non-air-entrained concrete. Properties of polymer
impregnated concrete which make it very attractive for use in concrete
hydraulic structures include low permeability to water and chloride pene
tration, excellent long-term resistance to freezing and thawing, and com
pressive and flexural strengths three to four times greater than those of
ordinary concrete.
26
3. Addition of Reinforcement. As previously mentioned, cracked
reinforced-concrete members have been successfully repaired and upgraded
by adding additional reinforcement to the member either internally or ex
ternally. The addition of internal reinforcement, referred to as post
reinforcement, has been used successfnlly for a number of years by the
Kansas Department of Transportation, which developed the process to
repair cracked bridge deck beams and girders. As a result, much of the
necessary materials, equipment, and expertise necessary for the
successful application of the repair technique already exists. However,
since apparently all applications of the technique have been limited to
bridge deck beams and girders, it may be necessary to modify or redesign
some of the equipment in order to use it with other types of structures.
External reinforcement, i.e., the addition of steel rods, plates,
and reinforcing tendons to the exterior surface of a member, has primari
ly served as a means to upgrade the strength of an under-reinforced or
highly cracked concrete member, although it can be used as a means to
pull cracks closed. The major limitation regarding this form of repair
is that strengthening and stiffening the member being repaired may lead
to cracking in other parts of the structure.
Both methods, however, appear to offer one of the best means of re
pair when it is necessary not only to seal cracks but also to restore or
upgrade the structural characteristics of a deteriorated member.
4. Thin Reinforced Overlays and Shotcrete. Once the cracks in a
member have been repaired, it may be necessary to resurface the face of
the member to cover up any scars or imperfections left by the repair
procedure, to repair minor damage resulting from spalling, or to provide
the member with a more durable wearing surface. Thin reinforced over
lays, either bonded or unhanded, and shotcrete made using either conven
tional portland-cement mortar or concrete-polymer materials appear to
offer the best approach. It should be pointed out that isolated patching
of the surface may be necessary before renewing the surface, if damage
due to cracking and spalling is excessive.
27
Both methods, overlaying and shotcreting, offer a very practical
means for cosmetically renewing the surface of a member once it has been
repaired. It may be necessary, however, to use reinforcing in both re
pair procedures to reduce the possibility of reflective cracking in the
new surface. The use of concrete-polymer materials offers a means of
providing surfaces with superior durability characteristics as well as of
reducing the ability of water to penetrate into the member once it has
been repaired.
The equipment and expertise required for shotcreting already exist,
although some work may be necessary in these areas if concrete-polymer
materials are to be shotcreted. Work in these areas may also be neces
sary in order to optimize the placement of overlays to vertical surfaces.
28
PART III: CASE HISTORIES
Presented in this section are case histories illustrating in situ
applications of each of the three repair techniques recommended in the
preceding section for use in the repair of cracked concrete hydraulic
structures.
Pressure Injection
As mentioned earlier, pressure injection, primarily with epoxy re
sins, has been used extensively for about 25 years to repair various con
crete hydraulic structures. Examples of hydraulic structures repaired
with this technique include the Daniel Johnson Dam in Northern Quebec and
the Pacoima and Twin Lake Dams in California. Pressure injection has
also been tested, on a small scale, by the Rock Island District of the
U.S. Army Corps of Engineers to repair concrete pier stems in Lock and
Dam No. 20 on the Mississippi River, Canton, Missouri.
1. Daniel Johnson Dam. In 1981, massive cracks in the main arch
buttress and on the face of the first arch to the left of the central
arch were repaired using epoxy injection (Adhesives Engineering Bulletin,
1981a). The cracks were discovered during construction on a turbine
addition to the structure. Located on the Manicaughan River in Northern
Quebec, the Daniel Johnson Dam is a 700-ft-tall multi-arch dam with a
reservoir covering 800 sq mi of water.
An inspection of the dam indicated that two major fissures needed to
be repaired. The first was a shallow 4,000-ft2 fissure intersecting the
face and penetrating to a depth of approximately 5 ft, and the second was
a 7,000-ft2 crack ranging from 7 to 12 ft in depth. The shallow fissure
was generally wider than 1/4 in., while the deeper crack was less than
that.
After the damage was inspected and evaluated, a repair system was
developed which called for the grouting of the cracks that were >1/4 in.
with a cement slurry and injecting the cracks <1/4 in. with a structural
epoxy adhesive.
29
Work began by drilling l-in. diam holes on 5-ft centers to intercept
the shallow fissure. Steel anchor rods were grouted into some of these
holes wh~}e others were used as grouting ports. The surface of the fis
sure was sealc·'l with a cement pas L c. • Cement grout was then pmnped into
the fissure to fill all the large voids. After the grouting was complet
ed, cores were taken, and any small fissures which were found to be empty
were grouted with epoxy adhesive.
The deeper and narrower fissure was repaired in a similar manner,
using the epoxy adhesive. However, dyed water was first injected into
the crack to determine the communication pattern and the volume of
adhesive needed to fill the crack. Upon completion of the dye test,
the cracks were pressure injected with the epoxy adhesive. A total
of approximately 1000 gal of Adhesive Engineering's Concresive 1380
injection adhesive epoxy was used to seal the cracks.
2. Pacoima Dam. In 1972, over 470 lin. ft of cracks were repaired
in the walls of the spillway tunnel of the Pacoima Dam, located about
4 mi northeast of San Fernando, California, in the Pacoima Creek
(Adhesives Engineering Bulletin, 1973).
The reinforced concrete dam, completed in 1929, has a crest height
of 365 ft above the original streambed, a crest length of 640 ft and a
base thickness of 100 ft. The 300 ft long spillway tunnel, cut through
a mountain, has two inlets. The tunnel is 15 ft in diameter and has a
12-in. thick reinforced concrete wall.
The smaller cracks in the tunnel walls were caused by aging and the
temperature extremes in the tunnel, and the larger cracks by the severe
shocks of the February 1971 earthquake in the San Fernando Valley. The
cracks varied in thickness from hairline to 1/4 in. Water seepage
through the cracks and construction joints had discolored the concrete
and created a buildup of efflorescence.
The damage was repaired by first thoroughly wire brushing the cracks
and construction joints to remove all traces of dirt and loose foreign
material. Where required, both temporary seals and permanent epoxy
30
adhesive seals were placed over the cracks. Injection ports were placed
approximately 6 in. on center along the sealed cracks. Adhesive Engine
ering's Concresive 1050 epoxy adhesive was then injected at moderate
pressure from port to port until all cracks were penetrated fully and
sealed.
During the repair work, extreme weather conditions were encountered,
including below-freezing ambient temperatures. Space heaters brought
into the tunnel to help raise the ambient temperatures did not help very
much. Nevertheless, the epoxy cured out completely, although curing took
longer than usual.
Cores removed from the repaired tunnel walls revealed complete pene
tration of the cracked concrete, even to the finest branching cracks.
3. Twin Lakes Dam. Constructed in 1923, the Twin Lakes Dam is a
reinforced concrete arch structure located on Caples Lake near Carson
Pass in the California Sierras (Adhesives Engineering Bulletin, 1971).
The dam is about 30 ft high and 70 ft wide. The main arch is 18 in.
thick at the bottom, tapering to 10 in. at the top.
Vertical cracks at five different locations in the face of the arch
section totaled approximately 120 lin. ft and ranged from hairline up to
a maximum of 1/8 in. in width. Water was leaking through some of the
cracks. In addition, the concrete had begun to spall around the cracks
and efflorescence was present in much of the area to be repaired.
In some areas, conventional patching materials used in the past to
repair 1/16-in.-wide dry cracks had disbanded and spalled off. The worst
leak in the main arch developed in one large diagonal crack which had
been surface grouted.
To repair the concrete arch, the reservoir was first dewatered to
allow access to the spillway from both front and back sides. The con
crete surfaces were then cleaned thoroughly to remove loose concrete,
dirt, and efflorescence, and the cracks and adjacent concrete were washed
down with a solvent.
31
A surface seal was applied over the cracks on the front and back
faces. Steel ports were inserted at intervals along the cracks. Adhe
sive Engineering's Concresive 1050 epoxy adhesive was then pressure
injected through the ports for the full depth of the cracks.
Because the repair work was performed in September and October, R
plastic film structure was erected around each of the five areas to be
repaired to help shield the work areas from the extremely cold tempera
tures normally encountered in the Sierras during these months. Ambient
temperatures within the enclosures were 65°F to 700F. This assured
optimum conditions for rapid cure of the epoxy adhesive.
4. Lock and Dam 20, Pier 39, Canton, Missouri. One of the best
documented applications of the epoxy pressure-injection repair technique
found was the work done by the Rock Island District of the U.S. Army
Corps of Engineers (Flock and Walleser, 1983). In August 1982, Pier 39
of the Lock and Dam 20 on the Mississippi River, Canton, Missouri, was
repaired using epoxy injection. The work was primarily done to evaluate
the applicability of the repair procedure.
Construction of Lock and Dam 20, using non-air entrained concrete,
was completed in 1935. Periodic inspections of the structure over the
years indicated extensive cracking in many of the concrete pier stems
supporting the dam service bridge. Cracking was primarily attributed to
structural stresses. This cracking, however, was apparently aggravated
by freezing and thawing due to inadequate drainage on the pier tops which
allowed rainwater to pond on the horizontal tops of the piers and infil
trate down through the piers.
As a result of the apparent continuing deterioration of the con
crete, a private contractor performed a prototype test to investigate the
use of epoxy injection as an alternative to conventional repair methods
of removal of unsound concrete and replacement with air-entrained con
crete. The proposal submitted by the contractor called for drilling
1-in.-diam ports in areas of severe damage. Water would be pumped into
the ports under pressure to measure flow quantities, trace flow patterns,
and estimate the severity of the damage. Cracks on the exterior of the
32
pier from which water flowed during the test would be surface sealed pri-\
or to injection. Following surface sealing of the cracks, a grid of
1-in.-diam injection ports \i~l.Jld be drilled in areas of high water flow.
A two-component low-viscosity epoxy adhesive, Adhesive Engineering's
Concresive 1380, would be injected into the ports through packers. The
proposal indicated that cracks 0.002 in. or wider would be sealed.
Before beginning repair operations, ultrasonic pulse velocity mea
surements were taken through the concrete in the pier. It was anticipat
ed that the comparison of data obtained by taking sonic readings before
and after epoxy injection would be useful in evaluating the effectiveness
of the epoxy repair.
Results of the preliminary pressurized water test indicated the ex
istence of a more extensive crack network than was originally anticipat
ed. As a result, the original work proposal was modified to require the
exterior surface of the pier to be completely sealed in order to prevent
leakage of the epoxy at the pressures required for injection.
Before the exterior surface of the pier was sealed, lime deposits
and other loose material were removed from the pier by a portable
sandblast unit. As these deposits were removed, water from the
preliminary pressure test continued to seep from various cracks,
especially inside the archway.
Several epoxies were used to seal the exterior of the pier. Epox
ies of thick consistency were applied by trowel to the upstream and down
stream faces. These were the areas which exhibited the greatest amount
of surface cracking. Thinner epoxies were applied to the ceiling of the
archways, walls of the access manhole, exterior walls, and the top of the
pier. Silica sand was broadcast into the epoxy applied to the top of the
pier to provide skid resistance.
After the epoxies used to seal the pier had cured, additional pres
surized water tests were performed at a maximum water pressure of 40 psi.
Several small leaks were detected in the surface seal. These areas were
resealed prior to epoxy injection.
33
Epoxy injection was started in the middle of the archway ceiling at
Port No. 9 (refer to Figures 10-12). Pumping was continued, while the
epoxy reached injection Por~ s Nos. 2, 3, and 12 on the downstream face.
The operation was then moved to the upstremn ceiling port (No. 8). The
contractor had difficulty maintaining a seal around the inJection packer.
In addition, leakage of epoxy was noted at the bottom edge of the access
manhole and adjacent to the electrical conduits embedded in the archway
ceiling. An attempt was made to inject epoxy into Port No. 10; however,
it was impossible, even at a pressure of 160 psi.
EL, 501.0
I I I I I I 1 8 9 10
____ J ___ :l __ l __ l __
INTO CEILING OF ARCHWAY,
EL, 4~"'""'-'=0 __ '-----------::----,,.....-------\
V -CONSTRUCTION JOINT _)
Figure 10. East elevation showing location of injection ports (Flock and Walleser, 1983).
34
\.i.J VI
EL. 501.0
• llf
V - CONSTRUCTION EL, 4~.0 JOINT ~
'-----1---...L....--...J EL, 486,75
Figure 11. Downstream view showing major cracks and locations of injection ports (Flock and Walleser, 1983).
UPSTREAM
5\.- /4
7 ' \6
DOWNSTREAfY\
Figure 12. Top view showing major cracks and locations of injection ports (Flock and Walleser, 1983).
The injection operation was then moved to Port No. 12 on the down
stream face. Again, problems were encountered with leakage at the packer
seal. Injection was moved to Port No. 11. Leaks through the seal coat
developed at pop-outs and along wider cracks on the downstream fC~c.e. As
epoxy v:as pumped into the pier, water was driven from the cracks and
appeared on the surface of the pier at elevations below the seal coat.
Epoxy also appeared at several small cracks in these areas. Injection
was continued at Port No. 13, then at Port No. 1. Epoxy ponded at the
top of the downstream east grout pad while pumping at Port No. 1. Injec
tion was continued at Port No. 6 on the pier top. The last port injected
was No. 7, also on top. This port was injected until epoxy appeared at
the handhold at the top of the access manhole. The injection packers
were then removed and all ports were sealed with epoxy. The pier was in
jected for a total time of approximately 34 hours. Approximately 8.4 gal
of epoxy were injected into the pier.
Ultrasonic pulse velocities taken after the pier had been injected
showed that some of the stations exhibited an increase in velocity over
the original values. Higher velocities generally indicate sound con
crete. The most significant increases occurred at locations on the lower
part of the pier. Again, consistent readings could not be obtained
around the top of the pier, indicating greater deterioration in the
concrete in the upper portion of the pier stem.
Upon completion of the pulse velocity tests, concrete cores were
taken from various locations to further evaluate the effectiveness of the
repair. Presented in Figure 13 are photographs of cores taken before and
after the pier was repaired. Cores removed from the pier top contained
numerous random and crudely parallel diagonal cracks as well as many
fractured aggregate particles. Only 5 to 10% of the cracks were filled
with epoxy, explaining the difficulty in obtaining pulse velocity
readings.
36
VJ "---
LO PIER 39
Figure 13. Cores taken from Pier 39 before and after epoxy injection.
The cores taken from the downstream face were not as severely
cracked as the core removed from the pier top. The cracks were finer
and more randomly distributed throughout the length of the cores. Fifty
to ninety--five percent of Lhe cracks were filled with epoxy. Tests for
resistance to freezing and thawing subsequently run on two of these cores
indicated the epoxy injection was successful in rebonding the concrete
and in holding it together under freezing and thawing conditions. Most
of the failure observed in these tests resulted from crumbling of the
non-air entrained mortar and cracking of the coarse aggregate particles.
In general, it was concluded that the epoxy in the fractures tended to
hold the concrete together during the freezing and thawing cycles and
aided in slowing down the deterioration of the concrete. Based on the
tests, the performance of the concrete in the structure cannot be
predicted absolutely; however, it is expected that while the exposed
concrete surface would be subjected to damage due to freezing and
thawing, the deeper interior concrete may be cemented together well
enough to form an integral whole and to perform satisfactorily.
An inspection of the repaired pier stem, after one winter of ser
vice, however, indicated cracks in many areas of the epoxy surface coat
ing on both the upstream and downstream faces of the pier as well as on
the pier top. Leaching and calcium deposits were also present on the
epoxy-treated surfaces as well as on some untreated surfaces. It was
subsequently concluded that the deterioration observed in the pier stem
was primarily due to continued lack of adequate drainage of rainwater
from the top of the pier stem and a failure to adequately seal the pier
top. Thus, water was still penetrating into and through the concrete, as
evidenced by the calcium carbonate deposits that appeared after the com
pletion of the epoxy injection work. In addition, it was concluded that
the epoxies used to coat the pier were much too brittle, as evidenced by
the reflective cracking.
From the experience gained at Lock and Dam 20, Flock and Walleser
(1983) made the following recommendations regarding any future epoxy
injection work.
38
(1) A visual survey of the damaged concrete should be made before
locations for installing ports are selected. It is generally recommended
that ports should be spaced at 1 to 1-1/2 times the depth of a crack.
However, if a crack extends through to both sides, a grid of ports should
be drilled to a depth of 1 ft 3 in. from the outside of the piers and
staggered with ports drilled to the same depth from inside the pier
arches.
(2) Water pressure tests should be made separately from the epoxy
injection contract to estimate the severity of damage and to determine
the location of ports.
(3) A phase-injection, multi-pump system should be considered.
This system would consist of drilling grids of ports to different depths.
(4) Sandblasting is generally thought to be the most effective
method of cleaning concrete surfaces. This belief proved to be true at
Pier No. 39. The contractor attempted to remove incrustation by water
jetting, but was unsuccessful.
(5) Wide cracks, such as those on the downstrerun face of Pier No.
39 where leakage occurred during injection, should be more positively
sealed. Consideration should be given to "V" routing of these cracks and
filling the "V" sections with an epoxy mortar.
(6) Injection packers must also be firmly sealed to withstand hy
draulic pressure. The contractor used removable packers at Pier No. 39.
A better method would be to use permanent-type packers which could be
embedded deeper into port holes and cut off flush with the concrete after
injection.
(7) In general, it would appear advisable not to use a filled epoxy
for sealing the surfaces of the piers. If a rougher surface texture is
required for paint, the epoxy seal can be scarified or even removed by
heating to above 572°F.
(8) Provision should be made for repairing spalls and popouts. A
filled epoxy should be used.
39
(9) The epoxy "pushed" or displaced water as it was injected into
the pier. More positive relief could be provided by ensuring that the
proper number and grid pattern of ports are selected prior to epoxy
injection.
(10) Guideline specifications provided by epoxy resin formulators
are basically applicable only to well-defined, two-dimensional cracks.
The criterion for evaluating a satisfactory repair is that a crack be 90%
filled. There is no direct correlation of this criterion to the three
dimensional, discontinuous type of cracking which is occurring in piers
at Dam 20.
(ll) Proper mixing, surface preparation, application, and selection
of epoxy formulations are essential for good performance of concrete re
pairs. Furthermore, the repair at Dam No. 20 is not entirely applicable
to present formulation guideline specifications. Therefore, reputable
and experienced contractors familiar with current techniques should be
employed.
5. Equipment, Labor, and Economics. The major piece of equipment
required in pressure injection work is the pumping system used to inject
the adhesive into the cracks. The injection process normally uses two
positive displacement pumps geared together to provide the proper ratio
for the adhesive components, an electric or air motor drive for the
pumps, and a static mixing head. The exit nozzle of the mixing head is
held against the face of the crack at entry ports which have been created
by leaving an interruption in the surface-applied seal. For uniform sur
faces and cracks which require relatively low injection pressures, 100
psi, a hot-melt thermoplastic seal can be used. In other cases, a
cementitious, polyester, or epoxy seal is used. For cracks requiring
very high pressures, 200 to 300 psi, to achieve penetration, a pipe
fitting may be bonded into a hole which has been drilled with a hollow
core drill to intersect the crack.
Other pieces of equipment may include a sandblasting unit for clean
ing the exterior surface prior to sealing, a spray application unit for
applying a seal coat to exterior surfaces, drills for inserting the in
jection ports, and assorted mixing and clean up supplies.
40
Labor required for any given job will be totally dependent upon the
size and magnitude of the work to be done. Small jobs can probably be
handled by a work crew of four or five men, whereas large jobs may re
quire two or three times that number.
A unit cost for epoxy injection work is difficult to determine, since
each job is bid individually on the basis of the type of structure to be
repaired and the unique problems associated with that job. Material
costs can be roughly determined by multiplying the number of linear feet
of crack to be repaired by a factor of $2 to $4, depending upon the width
of the crack to be repaired. This figure can then be doubled to obtain
an estimate for the cost of labor.
Polymer Impregnation
Polymer impregnation has been used to rehabilitate a variety of
portland-cement concrete structures including a structural floor slab
in the Cass County Jail, Fargo, North Dakota; the deck of the Greenport
Bridge, Greenport, New York; and an outlet tunnel wall and part of the
stilling basin floor at the Dworshak Dam, Orofino, Idaho.
1. Cass County Jail, Fargo, North Dakota (1976) (Kaeding, 1976 and
1978). Built in 1913, the three-story reinforced Cass County Jail was
condemned in 1974 because of its severe deterioration. Examination of
the structure had revealed extensive cracking in the beams, columns, and
walls as well as pockets of soft and weak concrete; layers of laitance
within the walls; chemically induced deterioration; areas of segregation;
weak, unwashed, and ungraded aggregates; and foreign material in the
1natrix. A load test, however, verified that except for the attic floor
slab, the slabs on the occupied floors met the load capacity criteria of
the American Concrete Institute's (ACI) building code requirements for
reinforced concrete.
The 5-in. attic slab, which was an integral part of the structural
frame, exhibited much more severe deterioration than the rest of the
structure. Attempts to remove cores from the attic slab failed when the
concrete disintegrated in the coring machine. This slab was repaired
41
using polymer impregnation after other methods of repair were judged un
suitable. Removal and replacement of the slab were not attempted because
of the overall weakened condition of the remainder of the structure,
which would have creatPd the danger of total collapse of the stru...:ture
during L:1rge-scale demoliti0r~
The polymer impregnation process usually consists of four steps:
(1) drying the concrete to remove free moisture present in the pore
system, (2) cooling the concrete after drying, (3) impregnation of the
concrete with a low-viscosity liquid monomer system, and (4) in situ
polymerization of the monomer.
Because of the highly porous condition of the concrete in the slab, a
void volume of about 27%, and low moisture content of 1.6%, it was un
necessary to dry the slab prior to impregnation. In fact, a preliminary,
small-scale impregnation test on the slab indicated that the monomer
penetrated the concrete so easily that it would be necessary to coat the
bottom of the slab with a membrane to prevent the monomer from leaking
out of the bottom of the slab.
The procedure used to impregnate the attic slab was as follows. The
slab's surface was first swept clean to remove any dirt and debris. A
1/4-in. sand layer was then placed uniformly over the slab to hold the
monomer in place during the soaking period. A pipe manifold system with
spray nozzles was then constructed over the top of the sand layer for
use in spreading the monomer over the sand blanket. A sheet of polyeth
ylene was then placed over the area to be impregnated, to provide a vapor
control cover. While this work was being performed, the bottom of the
slab was sandblasted and then coated with an epoxy compound to provide an
impervious membrane.
A monomer system consisting of 89.5 wt% methyl methacrylate (MMA) -
10 wt% trimethylolpropane trimethacrylate (TMPTMA) - 0.5 wt% azobis iso
butyronitrile (AIBN) was used to impregnate the slab. The monomer system
was applied to the sand blanket in four successive passes, with enough
time between each pass for the monomer to soak into the concrete. A
total of 3200 gal of monomer was used to impregnate the 7096-ft2 attic
slab at an impregnation rate of 0.125 gal/ft2 per hour.
42
Once the slab had been impregnated, the sand blanket was removed and
electric resistance heating elements were placed above the surface of the
slab. The slab was then heated 1or 18 hours to polymerize the monomer.
The temperature &t the top surface of the slab was maintained at 170°F,
while the temperature at the bottom surface of the slab was held at
145°F.
Upon completion of the polymerization cycle, the slab was cored and
tested to determine the change in compressive strength. A final compres
sive strength of approximately 3000 psi was obtained for the impregnated
concrete as opposed to an initial compressive strength of less than 800
psi.
In addition to rehabilitation of the attic floor slab, a number of
other repairs were carried out to rehabilitate other structural members
for a total cost of $887,343. The polymer impregnation work cost ap
proximately $180,000. Had it been possible to remove and replace the
attic slab, the estimated cost was about $42,000. The only alternative
to restoring the structure was to demolish and replace it at an estimated
cost of $4 million.
2. Greenport Bridge, Greenport, New York (1977-78). Polymer impreg
nation was used to restore the structural integrity of the badly deterio
rated concrete deck of the Greenport Bridge, located on Route 25 between
the villages of Greenport and Southold, Long Island, New York (Fontana
and Kukacka, 1979).
The Greenport Bridge is a skewed, two-lane, three-span through girder
with transverse floor beams and a reinforced concrete deck structure,
built in 1929. The 3800-ft2 bridge deck was originally a two-course con
struction deck with an 8-in. structural slab and a 4- to 6-in. wearing
course separated by a bituminous membrane. At the time of rehabilita
tion, the deck was so badly deteriorated that it was impossible to remove
an intact core.
43
The deck was impregnated one lane at a time, with each lane being
divided into three sections. In general, the procedure used to impreg
nate each section was as follows. The wearing course and bituminous mem
brane were first removed and replaced with a temporary steel deck at
roadway grade so that traffic could be maintained during impregnation of
the structural slab. The steel plates used for the temporary deck were
provided with a skid-resistant surface composed of silica sand bound with
a polyester resin. Once the temporary steel deck was in place, the
structural slab was dried by electric infrared heaters located beneath
the deck. An enclosure consisting of semirigid fiberglass insulating
board backed with aluminum foil was installed directly beneath the heat
ers to reduce heat loss to the structural slab.
Before the deck was dried, the underside of the bridge slab and the
floor-beam and girder encasement concrete were sandblasted. Once the
drying cycle was complete and the deck was allowed to cool, the underside
of the bridge slab was coated with a polyester seal coat. The coating
was done after the drying cycle was completed and before the impregnation
cycle was started, to prevent thermal degradation of the sealant
material.
The deck was then impregnated with a monomer system of 95 wt% methyl
methacrylate (MMA) - 5 wt% trimethylolpropane trimethacrylate (TMPTMA),
to which 0.5 wt% initiator azobis isobutyronitrile (AIBN) was added. The
monomer was applied to the deck through a distribution system located
between the temporary steel deck and the graded aggregate which had been
placed over the structural slab once the wearing course was removed.
The rate of monomer application was dependent on the rate of absorp
tion of monomer into the deck. The total impregnation time for each sec
tion was three to four days. Once the deck was saturated, the monomer
was polymerized by reheating the deck using the infrared heaters.
44
After impregnation of the structural slab, the temporary steel deck
was removed and wearing courses were placed to roadway grade, with poly
mer concrete used in the westbound lane and an asphaltic concrete in
the eastbound lane. Several difficulties arose during the first drying,
impregnation, and curing cycles. The temporary deck was not watertight,
and water carried onto the deck by traveling vehicles seeped through the
plates and resaturated the concrete. This resaturation substantially
prolonged the drying cycle. During the first impregnation cycle, the
monomer leaked through the polyester seal coat, which was ineffectual in
retaining the monomer in the deck. The leaking monomer and some possible
malfunctioning electrical component caused a monomer fire during the
first curing cycle.
A reevaluation of the construction procedures delayed the project for
several months until revised procedures could be instituted. The new
procedures included use of a flexible silicone rubber coating to replace
the brittle polyester coating and use of a gaseous carbon dioxide (COz)
suppression system to eliminate the possibility of a monomer fire. The
silicone rubber coating greatly improved the retention of the monomer in
the structural slab, but because of irregularities in the bottom surface
of the slab, it did not form a continuous film, and some monomer leakage
was still evident. The COz suppression system worked very well to elimi
nate fires on top of the deck. The remaining drying, impregnation, and
curing cycles proceeded without major problems.
Once the deck was fully impregnated, full-depth cores were removed to
evaluate the success of the repairs. As stipulated in the contract spe
cifications, the impregnated structural slab was to be restored to the
concrete slab's original design strength of 3000 psi in compression or
higher. Results of the compressive strength tests indicated that the
polymer-impregnated concrete (PIC) had an average compressive strength of
4320 psi.
The polymer loading in the deck, measured by thermogravimetric analy
sis, was found to vary between 3.1 and 17.5 wt%. This large variation
can be attributed to differences in the quality of the concrete through
out the deck, not to differences in the impregnation procedure.
45
The water absorption of the PIC varied between 0.5 and 1.5 wt%. Two
cores were subjected to 50 cycles of freezing and thawing, as per ASTM
C-666, with no appnrent loss of weight.
In general, all cores were well bonded ar1d the polymer seemed to be
completely dispersed throughout the core.
When evaluating the costs of the Greenport project, it must be
remembered that the purpose of this experimental project was to test the
feasibility of the proposed construction procedures for reconstituting
deteriorated bridge decks in high density traffic areas by using polymer
impregnation and to evaluate the long-term performance characteristics
and the cost effectiveness of such repair. The actual cost of the
impregnation work was $809,670. This was $40,404 over the contractor's
original bid price. The increase in cost was due to a change in orders
after the project had been started. The original engineer's cost esti
mate was $585,547. For this project, the rehabilitation of the struc
tural slab by impregnation was 70% higher than the estimated cost of
$475,000 to replace the structural slab by conventional means. However,
for a comparable project in a high-traffic area such as New York City,
the costs of the two methods would probably be much closer together.
Building a full-service, temporary structure would necessitate condemning
buildings, acquiring of property, and constructing a full-size structure.
Since basic impregnation costs should not increase at a comparable rate
for this much larger project, rehabilitation by impregnation would prob
ably be cost effective in urban areas.
3. Dworshak Dam, Orofino, Idaho (1975). Polymer impregnation tech
niques were used to repair major cavitation/erosion damage iri a portion
of the stilling basin floor and the walls of an outlet conduit of the
Dworshak Dam (Murray and Schultheis, 1977; Schrader, 1978; and Schrader
and Kaden, 1976).
The Dworshak Dam is a straight gravity structure, 717 ft high, with a
crest width of 3287 ft. The dam contains three similar regulating out
lets that are 12 ft wide by 17 ft high.
46
The outlets, first used in 1971, had been used intermittently for a
total of 10 months before being repaired. The outlets were inspected
in June 1973, at which time some minor, isolated surface scaling was
noticed. A year later, two outlets showed severe cavitation damage 50
to 75 ft downstream of the outlet gate. Severe damage was described as
massive removal of concrete and removal of reinforcing steel. The worst
area had a depth of concrete failure of approximately 22 in., and some of
the No. 9 reinforcing bars were missing. Downstream of the severely
damaged areas in Outlet No. 1, medium damage, less than 1 in. in depth,
totaling over 60 yd2 of surface area was found throughout the remaining
200 ft of the outlet. Surrounding these areas of medium damage were
areas of light erosion or surface scaling as well as areas where the
original curing compound was still intact. Every horizontal construction
joint in the outlet showed typical scaling and the beginning of failure.
The stilling basin received severe damage which removed approximately
1600 yd3 of high-strength reinforced concrete from its 29,261 ft2 floor
surface. The damage was primarily attributed to debris and gravel trap
ped in the stilling basin. Damage varied from 0.1 to 9 ft in depth.
The procedure used to repair the walls in Outlet No. 1 were as
follows. Severely damaged areas were repaired by outlining them with a
3-in.-deep saw cut and then removing all the damaged concrete within the
saw cut to a minimum depth of 15 in. Reinforcing bars were reset within
the chipped out areas, which were then filled using fibrous concrete.
Areas of medium damage were bush hammered back to a depth of 3/8 to
1 in. to sound concrete and then the voids were filled with "dry packed"
concrete mortar. Once these repairs were completed, all wall surfaces
in Outlet No. 1 were impregnated to a height of 10 ft.
The concrete in the outlet tunnel walls was impregnated in 10- by
12-ft sections using the following procedure.
47
The concrete was dried using infrared heaters mounted inside 12- by
14-ft insulated plywood boxes. The boxes were placed against opposite
walls and braced against themselves during the drying operation. The
concrete was dried at a surface temperature of 275°F for a period of 5
hours, after which the concrete was allowed to cool to a su''f ace tempera
ture of 90°F. Once the wall had cooled down, it was impregnated with a
monomer system consisting of 95 wt% MMA - 5 wt% TMPTMA - 0.5 wt% AIBN.
Soaking was accomplished using two 10- by 10-ft stainless steel pan
els, which were pressed against the outlet walls to form shallow tanks to
contain the monomer. The panels were mounted on a cart, which was pulled
up the outlet by cables. The gasket around the perimeter of the panel
provided about a 1/8-in. gap between the wall and the panel. To assure
that this gap was maintained throughout the soak area, spacers were tack
welded to the panel surface. A bead of caulking was then placed around
the outside panel edge next to the gasket, and waterproof tape was placed
over that with half the tape on the concrete and half on the panel edge.
This seal, however, did not prevent leakage problems at construction
joints and open structural cracks. It was eventually determined that
leakage could be controlled by applying a coat of gel epoxy to the
concrete wall surface where the perimeter gasket made contact. When open
construction joints and wide cracks were encountered, they were injection
grouted with epoxy at the perimeter of the soaking panel.
The soaking panels were filled through a manifold at the bottom of
the stainless steel plate. The manifold was connected to monomer storage
tanks which were pressurized to force the monomer into the soaking cham
ber. A sightglass located at the top of each chamber was used to deter
mine when the chamber was filled. By maintaining pressure on the storage
tanks, monomer was continually added from the reservoir at the same rate
that it soaked into the concrete. The concrete was soaked for 6 hours.
This was sufficient time to impregnate the concrete to a minimum depth of
0.5 in.
48
After completion of the soaking period, the excess monomer was drain
ed from the chambers back into the storage tanks. The chambers were then
filled with hot wate1 to polymerize the monomer. Electrical resistance
heaters were installed on the back of the soaking chambers to elevate and
maintain the water temperature. The backs of the 'l::tnels were insulated
to reduce heat loss and to protect the workmen. As a safety precaution,
the heaters were not activated until the chambers were filled with water
and no explosive vapors were present. The water temperature w·as main
tained between 150°F and 210°F for a minimum of 2 hours. The polymeriza
tion procedure was then complete; the water was drained from the forms,
and the forms were moved. This procedure was repeated until the outlet
walls were impregnated.
During the time the impregnation work was being performed in the out
let tunnel, a vapor monitor with a remote sensor continually indicated
the atmospheric concentration of MMA vapor. A continuous strip chart
recording of these readings was maintained to give an accurate history of
vapor concentrations in the outlet at all times. l~en the vapor concen
tration reached 100 parts per million, a visual and an audio alarm were
activated, at which time personnel were required to put on protective
clothing and face masks, work was to stop, and steps were to be taken to
decrease the concentration. In actuality, the vapors generally stayed at
very low levels, in the range of 3 to 15 parts per million. The only ex
ception to this was during the first soak when a bad leak occurred at a
construction joint before the procedure for grouting those joints had
been established.
The erosion damage in the stilling basin floor was filled with a con
ventional concrete mixture to within 15 in. of the final floor surface.
This was topped out with a 15-in. overlay of fibrous reinforced concrete
mixture anchored with No. 8 bars. One half the floor was then surface
impregnated.
49
Concrete in the stilling basin was impregnated in 58- by 12-ft sec
tions. An enclosure with insulated plywood sides and a removable top was
placed over each area to be treated. Before the surface was dried, free
water was swept away or vacuumed up anr~ a bead of caulking vas placed
around the base of the enclosure to keep water from reentering during the
drying cycle. A 3/8-in. layer of sand was then spread over the surface
to act as a wick during soaking and to promote better distribution of the
monomer. Infrared heat lamps spaced at 24-in. centers and 18 in. above
the floor were installed in the enclosure, and a lightweight insulated
roof was placed over it. The lamps used a total of 93 kilowatts or 134
watts per square foot and normally accomplished the drying in 10 to 15
hours. At a thermostat setting of 450°F for the air inside the enclo
sure, a surface temperature of 220°F could be attained in about 2 hours.
With the thermostat at that setting, the surface stayed at 220°F to 320°F
for the remainder of the drying cycle. Moisture was allowed to escape
through small openings spaced about every 5 ft in the enclosure walls.
Temperatures at different depths within the concrete were measured during
the drying process of the first setup using embedded thermocouples that
had been cast into the fibrous concrete slab. After drying, the concrete
went through a controlled cooling process. The heat lamps were turned
off, but the roof sections for the enclosure were kept in place. After a
brief period, the roof panels were cracked open about 1 in. After a
total cooling period of about 6 hours, the surface temperature was about
120°F. The roof sections were then removed, and the monomer was applied.
The monomer system used to impregnate the floor of the stilling basin
consisted of 97.5 wt% MMA- 2.5 wt% TMPTMA, with 0.5 wt% AIBN initiator.
The monomer was applied with a 12-ft long sprinkler pipe, which was
moved back and forth above the surface to be impregnated until a pre
determined amount of monomer had been applied. The monomer flowed by
gravity from an elevated mixing drum, through a hose, and into the sprin
kler pipe. A polyethylene sheet was draped over the surface to reduce
the evaporation losses, the roof sections were reinstalled, and another
polyethylene sheet was laid over the entire enclosure to keep down any
escaping fumes. In the first application, a total of 55 gal of MMA was
50
applied. The dosage was gradually increased to about 80 gal per setup
(0.115 gal/ft2), with which the sand blanket would just begin to stick to
the fl(' 0 r, indicating that the concrete could not absorb any more mono
mer. Since application of the monomer began when the surface temperature
was near the point at which polymerization could start, the percentage of
cross-linking agent was decreased. This adjustment made the monomer
slightly more difficult to polymerize because the rate of reaction was
slower in the temperature range of about ll0°F to 140°F. However, since
very hot steam was used to polymerize the stilling basin floor, obtaining
complete polymerization was no problem. After a usual soak of 6 hours,
wet steam at 750°F was used to polymerize the concrete. The portable
steam generator used for this was fed by an air hose at 115 psi pressure.
The steam was distributed through a manifold in the center of the enclo
sure, and within a minute, the entire area being polymerized was uni
formly filled with steam. After 1-1/2 hours, the steam was turned off
and the process was completed. Total time for both the impregnation and
the polymerization cycle in the stilling basin was normally 24 hours,
excluding initial setup and final cleanup.
Upon completion of the impregnation work, cores were taken from both
the outlet tunnel walls and the stilling basin floor to evaluate the
effectiveness of the impregnation process and the depth of impregnation.
The cores typically indicated that impregnation depths varied between 3/4
and 1-1/4 in. for the outlet walls and 1/2 in. for the stilling basin
floor. Cores taken through the dry-pack patches in the outlet wall in
dicated that the polymer had penetrated completely through the dry pack
material and into the base concrete.
During impregnation of the stilling basin floor, moisture kept
migrating up through vertical construction joints and cracks. It was,
therefore, impossible to dry the concrete at these locations. In most
cases, the moisture evaporated as it came through the joint or crack, and
floor surfaces that were more than about 6 in. from them were adequately
dried. Consequently, cores taken at cracks and joints showed no polymer
impregnation, while cores away from the damp areas had been successfully
impregnated.
51
During the impregnation work, a number of random cracks were noticed
in the fibrous concrete floor. When the surrounding areas were dry, the
cracks showed a dark trail of moisture and became very apparent. These
cracks were not caused by the drying process. They were present in the
floor before drying but were not as noticeable until the surface was
cleaned and dried. After the polymerization cycle was completed, there
were no new cracks and there was no apparent growth of the old cracks.
If moisture could have been kept out of the cracks, they would have been
filled and structurally sealed with polymer.
The outlet walls absorbed monomer at a rate of about 0.3 gal/ft2,
regardless of whether the area was dry-pack patching, new fibrous
concrete, or the original conventional concrete. The impregnation depths
were also similar in each of the three materials. The stilling basin
floor absorbed all the monomer when it was applied at a rate of 0.086
gal/ft2. When monomer was applied at a rate of 0.115 gal/ft2, a small
amount of monomer remained at the surface. It is difficult to explain
why the outlet walls soaked in more monomer and correspondingly had
greater depths of impregnation than the stilling basin floor, but the
reason can undoubtedly be related to at least one of the following:
(1) Moisture was being forced up through the floor of the concrete
stilling basin by uplift pressure. This made drying more difficult
and refilled the microvoids with water between the depths of 1/2 to
1 in. during the cool-down cycle. During the soak phase, monomer could
not enter the concrete below 1/2 in., since it was already saturated with
water. (2) The cool-down period was not sufficient to allow concrete
1/2 in. deep or greater to reach the temperature where polymerization
would not be initiated. During the soak cycle, monomer would penetrate
to a depth of about 1/2 in., where temperatures were high enough to cause
polymerization. The polymerized monomer then acted as a barrier, which
would not allow unpolymerized monomer to soak into the concrete beyond
that depth. Since the amount of cross-linking agent was purposely
reduced to help avoid this situation, its occurrence is unlikely.
(3) The drying duration and/or temperature was insufficient to dry
52
concrete deeper than 1/2 in. For most of the drying time, the tempera
ture was probably 260°F to 300°F. Durations of drying for the stilling
basin were generally on the same order as for the outlets except that
some setups were a little longer and some were a little shorter.
The dam vms inspected several months after the work had been complet
ed and after the dam was back in service. The inspections indicated that
the PIC has held up very well, while some of the other repairs made to
the structure, such as epoxy mortar repairs, have not.
4. Equipment and Economics. As can be determined from the case his
tories cited above, a large inventory of specialized equipment is neces
sary to impregnate a concrete surface successfully. Although specific
equipment needs depend upon the nature and requirements of each individ
ual project, general needs include the following: an insulated drying
enclosure, heaters for drying the concrete, temperature recorders for
monitoring drying and curing temperatures, an impregnation enclosure
(required for vertical applications), monomer storage and distribution
equipment, monomer mixing equipment, caulking and sealing compounds, a
curing enclosure, heaters for curing the impregnated concrete, vapor
monitoring equipment, fire extinguishers, and safety equipment for indi
vidual workers, such as rubber gloves, boots, impervious aprons, and
respirators.
Monomer requirements depend upon the quality of the concrete to be
impregnated and the desired depth of impregnation. For good quality con
crete, impregnated horizontally, a monomer application rate of 1 gal/ft2
results in a depth of impregnation of approximately 0.5 in.
It is equally difficult to determine a unit cost for polymer impreg
nation work since each project is usually very different from the pre
vious one. A wide range of costs have been reported for impregnation,
varying from $7/ft2 for the work done at the Cass County Jail, to
$213/ft2 for the work done at the Greenport Bridge. This variation is
partly caused by the fact that contractors are unfamiliar with what may
be expected of them, and the wish to pay the cost of new equipment on a
single small job. However, costs should stabilize as contractors become
more familiar with the process and acquire the necessary equipment.
53
When polymer impregnation is being considered as an alternative to
other forms of repair, the cost and time lost in shutting down the facil
ity must also be considered. In many cases, impregnation work can be
scheduled and performed while a facility is being operated, with only the
immediate work area closed to operations.
Addition of Reinforcement
The addition of reinforcement to concrete structures has been used in
a variety of applications to help seal cracks and to strengthen under
reinforced structures. Presented below are case histories of four such
applications.
1. lohn Day Navigation Lock and Dam, Oregon (1981). The John Day
Lock and Dam is located on the Columbia River between Oregon and
Washington about 110 miles upstream from Portland, Oregon. The lock is
675 ft long, 86 ft wide, and provides a maximum lift of 113 ft (Figure
14). The lock began operation in 1968.
During an inspection in 1975, significant structural cracking and
related spalling were discovered in two of the lock monoliths, monoliths
17 and 19. By 1979, cracking and spalling had been detected in four
additional monoliths, monoliths 13, 15, 21, and 23. The structural
distress in the monoliths consisted of cracks that originated at and
propagated from the upper inside corner of the filling and emptying
culvert and terminated at the surface of the wall in the lock chamber.
Because of the continued progression of the cracking and spalling, a
series of remedial repair procedures were devised and implemented to halt
and correct the ongoing deterioration (Adhesive Engineering Bulletin,
1981b, Barlow, 1986, and Neuberger, 1982).
The objectives of the selected repair procedures were to restore the
existing cracked structure to near its original uncracked condition, to
eliminate the cause of the cracking, and to effect the repairs without
54
BASALT
WASHINGTON SHORE
17 MONOLITHS WITH CRACKS
A, PLAN VIEW OF NAVIGATION LOCK
~-- ~~-~:.~ ~' t
·:'!
RIVERSIDE (SOUTH) LOCK WALL
WASHINGTON SHORE
B, (ROSS-SECTION OF STRUCTURE (LOOKING DOWNSTREAM),
Figure 14. John Day Navigation Lock.
55
significantly affecting the operation of the facility. This was accom
plished using a three-phase repair program. The program consisted of:
(l) grouting the foundation rock, (2) installing rock anchors through
the cracks after injecting structural epoxy adhesive into the cracks and,
(3) repairiug the surface of the wall in the lock chamber (Figure 15).
A stress analysis using finite element models revealed two probable
causes for the cracking. One cause was due to the original procedure
used to fill and empty the culvert that, before being changed, produced
high hydraulic surge pressures in the culvert. The other cause was due
to the normal lock-full condition which produced excessive foundation
deformations because of a layer of weak-flow breccia rock. Both causes
produced high tensile stresses in the concrete and, because of inadequate
reinforcing steel in the cracked areas to distribute these stresses, led
to the ensuing cracking and spalling. In addition, the cyclic loading of
the lock walls, due to filling and emptying of the lock, contributed to
the continued propagation of the cracking and spalling.
In Phase One of the repair program cement grout was pumped into the
flow breccia sandwiched between the two layers of basalt to fill in any
voids or open joints within the rock mass. The cement grouting was done
in several stages beginning in 1980. Post-consolidation grouting deflec
tion measurements at the crest of the monoliths revealed that deflections
were reduced 50 to 60% from the pregrout condition.
In 1981, Phase Two of the repair program was performed. This phase
of the repair program involved the installation of 73 rock anchors and
the injection of structural epoxy adhesive into the crack network.
Finite element analysis had shown that the forces induced by the rock
bolt anchors would substantially reduce the tensile stresses in the
cracked areas at the top of the culvert.
Subsequent to the grouting of the foundation, 10-in. diameter rock
anchor holes were drilled from the outside of the monoliths into the
basalt rock below. This work was carried on without interrupting lock
56
BASALT
. . ----- . . /~·' . .,.
. ·. ( '-
• I . •. ,_./' • •
_/--.,(' / .. ./ "--. -.
. f FLOW BRECCIA
(AREA TO BE GROUTED)
~ . (
~~ ___ ).
/ ··) . - . .
. .-r---
~ . ": .,. . ). · ....... ., . , .·..... . . ,. .. ~' ·.· ·.
-~
' ....
~ ..
.. :·. ·~· .. ·
CULVERT
....
. ·"" • :c
• p
1
. . . ~ . ...
. ., . ...: . , · . ., -4 •• · .. .,. ,.
: .. . ~ ·. . ;.
~"·
. , . . .. . ..
• :4 : NORMAL 4. DOWNSTREAM
• ~ ~ 1----=._o:::.W;;:::A-T:O:-E~R~~.::::-EV_E ...... L-f_
·~;,:.J--·.~. · . OPENING /
: .. ' .. ,. ~ .. ... ·j . " ::~:.: ~.
Figure 15. Cross-section of lock wall, showing details of crack location
and supporting rock mass.
57
service. The holes were drilled at either 55° or 66° entry angles from
the horizontal. The allowable dev~ation to avoid drilling into either
the filling and emptying culvert or the lock chamber was 1 ft in 100 ft.
The holes varied between 138 and 172 f 1- ;-md terminated approximately
40 ft below the base of the lock.
The 73 rock anchors consisted of 37 seperate, 7-wire strands, 0.6
in. in diameter. The rock anchors consisted of four major elements: the
30-ft anchor ends, the inflatable grouting packers above the anchor ends,
the greased and polyethylene sheathed strands above the packers, and the
stressing end assemblies. Packers were used to assure a good grouting
job in the crucial anchor end area. Greased and sheathed tendons were
used to have the capability to retension or detension the rock anchors.
The rock anchors were generally spaced at 4 ft intervals without any
disruption to lock service. The 30-ft anchor ends were grouted during
slack times in lock usage.
Following the placement and grouting of the rock anchors, the lock
was shutdown for 30 days and dewatered so that the cracks in the mono
liths could be epoxy injected and the subsequent post-tensioning and
grouting of the anchors could take place.
In order to confine the cleaning fluids and ultimately the epoxy
adhesives, each of the monolith joints had to be isolated to stop migra
tion upwards and downwards of the cleaning solution. This was done
by drilling 6-in. diameter core holes from the lock wall through the
monolith's joints intersecting the culvert. The holes were positioned
above and below the primary and secondary cracks that intersected the
joints. The core holes were then backfilled with a rapid setting
cementitious, flowable, non-shrink grout to provide a plug.
The monolith joint faces were then sealed and ported for injection.
A 100% solids, epoxy paste adhesive was then injected into the joints,
with a single component piston pump, from the culvert until it vented
on the lock wall side. The consistency of the epoxy was such that it
penetrated the joint with a minimal amount of material migrating into
the cracks intersecting the joints.
58
To provide maximum penetration of the cleaning solution, approxi
mately 3300 lin. ft of 1-1/2 in. diameter entry ports were drilled into
the lock wall, approximately 4 to 6 ft on center, to intersect the
interior cracks. Once Lhe holes were drilled, they were flushed to
remove any remaining drilling slurry. Deformed bars with epoxy grout
tubes were then inserted into the holes. The bars acted both as a filler
to reduce the volume of epoxy injected into each hole and as reinforcing
to transfer stresses across the cracked section. The tubes acted as
filling and venting ports for the cleaning solutions and the epoxy
adhesive.
The entire crack length of 440 ft, extending throughout the six
monoliths in the culvert were then sealed and ported using an epoxy
adhesive. A bio-degradable alkaline-based detergent was then introduced
into the crack network through a manifold system to flush out river silts
and clay which can have a detrimental effect upon obtaining a good struc
tural bond. Incorporated into the cleaning solution was a dye to help
locate any leaks or breaks in the crack and joint seals. The cracks were
then flushed with clean water to remove the detergent and blown with air
to remove excess water and any remaining particles.
Epoxy injection of the cracks began in the culvert. A high strength,
creep resistant, rapid curing, low viscosity injection adhesive capable
of bonding in wet conditions was used (Adhesive Engineering's Concresive
1380). Injection continued until material began to vent on the lockside
wall vents. Injection then continued in the lock by injecting into the
vent that initially showed material and progressed up the 1-1/2 in.
diameter hole grid until the entire crack network was filled. Approxi
mately 600 gal of epoxy was injected into the crack network, completely
filling the primary and secondary cracks.
Two days after the injection was completed post-tensioning of the
rock anchors began. Each tendon was stressed and 24 hr later checked
again to insure proper tensioning. Coring from inside the lock chamber
into the crack network was performed by the Corps of Engineers to verify
filling of the crack network. The second phase of the repair program was
completed in September 1981.
59
In March 1982, the dewatered navigation lock was inspected to evalu
ate the effectiveness of the Phase Two repairs. This was the first
opportunity to inspect the repaired area since completion of the struc
tural repair work. The filling and emptying culvert was inspected
initially. A hairline crack was discovered in the epoxy patch at the
culvert north wall-ceiling interface. The hairline crack was most
evident in monoliths 21 and 23 where it was continuous throughout the
length of the monoliths. Traces of the hairline crack were also seen
in the other repaired monoliths (11, 13, 15, 17, and 19), but they did
not appear to be continuous and were evidently only in short lengths.
Because of no spalling or chipping at its edges, the hairline crack
showed no signs of working or being progressive in nature. After the
culvert was inspected, the lock chamber wall was inspected. Evidence of
recracking of the injected epoxy was found only in a monolith 15 ladder
well. The crack was hairline and did not appear to be progressive. No
evidence of recracking was found at any other areas and no traces of the
crack were seen at the epoxy patch areas where the previous cracking had
surfaced. Areas of previously drummy surface concrete were found to be
solid, indicating that the injected epoxy was bonding the cracked
surfaces together.
In general, the repairs were considered to be successful. Even
though there was evidence in the culvert of some hairline cracking of
the injected epoxy, the hairline cracking had not progagated through the
culvert wall and out to the lock chamber wall surface. In addition, the
nonworking of the newly discovered hairline crack edges indicated that
the 73 installed rock anchors were keeping the crack restricted and pre
venting any new large scale cracking and spalling. After 8 months of
lock use following the completion of the structural repair, the repaired
monoliths were in a stabilized condition and showed no signs of future
continued cracking and spalling. The total cost of the repair program
was approximately $3.5 million, a savings of approximately $900,000 over
other methods that had been considered.
60
2. Markland Locks and Dam, Ohio River (1981). Markland J~cks and
Dam is located on the Ohio River approximately mid-way between Louisville,
Kentucky, and Cincinnati, Ohio. The locks and dam were constructed in Lhc
late 1950's and edrly 1960's as a replacement of Lock and Dams 35, 36, 37,
38, and 39 which were constructed in the 1920's. Markland Locks and Dam
project consists of two 110-ft wide locks (600 and 1200 ft long), a dam
of twelve 100-ft wide tainter gates supported by concrete piers, and a
hydroelectric plant. A two-lane state highway bridge also crosses the
structure.
The locks which are situated on the left bank, are constructed of
concrete gravity walls founded on rock, and incorporate steel miter
gates. Filling and emptying of the lock chambers is accomplished through
longitudinal culverts located within the lock walls. Flow of water
through the culverts is controlled by tainter valves located in special
"valve monoliths" located near the upstream and downstream end of each
lock chamber. The tainter gate is located in the culvert at the bottom
of the recess. Raising and lowering of the tainter gate controls the
flow of water through the culvert and in the process, water rises in the
shaft and the recess. Depending on the function being performed and the
valve monolith involved, a differential head of up to 46 ft may exist
between the water level in the recess and the lower pool. When the lock
chamber is dewatered for maintenance work, the differential head may be
as much as 55 ft.
In the early 1960's, a number of longitudinal cracks began devel
oping in the valve monoliths. The cracks extended from the corners of
the main recess to smaller recesses or to the joints at the end of the
monolith. In addition, some cracks extended from the top of the lock
wall to the top of the culvert. With hydrostatic pressure being exerted
within the recess and only minimal transverse reinforcement in the ends
of the monolith, the danger existed that the monolith could fail by
splitting longitudinally. The risks and costs associated with such a
failure were considered unacceptable, so repair of the monoliths was
required (Keith, 1972).
61
Acceptable repair methods had to meet the following criteria. They
had to be structurally adequate, constructable, cause a minimum of damage
to the existing structure, provide for minimum disruption to river traf
fic during construction, leave no obstructuions to traffic after the job
was completed and be implementable at an acceptable cost.
The monoliths were repaired by drilling holes horizontally into each
end of the recess perpendicular to the face of the lock wall and install
ing high strength steel rods into the holes to resist the internal hydro
static pressures. Dywidag threaded bar rock bolts were used and were
anchored into place using Celtite polyester resin. The work was per
formed from a floating plant within the locks. This permited work at
various levels by simply raising or lowering the water level within the
locks. The work crew and barge were moved out of the lock at the end of
a shift to permit passage of traffic in the main lock.
A track mounted, diesel precussion drill, located on the barge, was
used to drill the holes into the valve monoliths. After the holes were
drilled and cleaned, cartridges containing the Celtite polyester resin
were inserted into the holes. Two types of cartridges were used.
Cartridges containing a quick-setting polyester were inserted within
the anchorage zone of the hole and cartridges containing a slow-setting
resin were inserted into the stressing zone. After the cartridges were
in place, the rock bolts were threaded into place using the drill. The
threading action activated the curing of the polyester resin by mixing
the initiator into the resin. Once in place and firmly anchored by the
polyester resin, the bolts were stressed to 0.7 ultimate strength using a
hydraulic jack and jacking chair. The bolts were held in place for a
minimum of 4 hr with anchoring nuts to assure that the slow setting resin
had fully cured. The nuts were then removed by flame cutting, and the
holes were filled with a non-shrink grout to form a surface flush with
the wall.
62
Work was started in August and completed in November 1981. A total
of 240 anchors were installed at a cost of $170,000. Although some prob
lems and difficulties WPre experienced b) the contractor, the repairs
were considered to be quLte acceptable.
3. Apartment Building, Brussels, Belgium (1982). A gas explosion
caused by human error took place in one apartment on the tenth floor of
a 26-story apartment building in Brussels, Belgium. The apartment was
almost completely enclosed by reinforced concrete walls which had been
designed to take the wind loading of the structure; therefore, almost all
the damage to the structure was limited to the apartment. Within the
apartment, however, the damage was considerable.
The inner walls and glass panels were blown away by the force of the
explosion. During the first phase of the explosion, the ceiling slab was
loaded upward by the overpressure; during the next phase, it was loaded
downward by the underpressure. As a result, the slab exhibited large
deflections, with a maximum deflection of 2 in. In addition, the con
crete at the surface of the ceiling slab was severely deteriorated, by
a fire which arose following the explosion.
After a careful examination, it was decided that the damage could be
repaired by removing the loose and unsound concrete and exposing a clean
sound concrete surface. These areas would be repaired using an epoxy
mortar, EPISOL-EMT (NV. Resiplast-Belgium). Steel plates would then be
bonded to the underside of the slab to reinforce it and to eliminate the
deflections caused by the explosion (Van Gernert and Maesschalck, 1983).
The deflections were eliminated using eight hydraulic jacks, mounted
on a special supporting structure. Next to each jack was placed an ad
justable screw. During and after the lifting operation these screws
served to secure the slab in the lifted position. With the hydraulic
jacks, the slab was put back to a nearly horizontal position.
At lifting, new cracks appeared at the top side of the slab. To pre
vent them from closing again when the supporting structure was removed,
they were filled with a low-viscosity epoxy resin. At the edges, cracks
63
originated at the bottom face of the plate. These cracks were filled by
pressure injection of epoxy resin. The steel plates were then glued to
the bottom face of the slab with epoxy, EPICOL-U.
Tlw steel plates had a cross section of 0.2 in. by 9.84 in. and were
spaced at 31.5 in. The cross-sectional area of these plates was deter
mined by the limitation of the deflections. Plastic deformations during
the explosion disturbed the continuity of the slab over its supports so
that deformability of the slab had increased. External reinforcement was
applied, which increased the stiffness of the slab, so that the deflec
tions remained within the limits allowed.
Special attention was paid to the design of the anchorage lengths of
the plates. The maximum shear stress in the epoxy joint was, therefore,
calculated. The maximum shear stress allowable in the joint corresponded
to the surface tensile stress of the concrete. The surface tensile
stress was measured by tear-off tests on small steel cylinders, glued
to the concrete surface.
After a curing time of seven days, the epoxy glue attained its final
strength, and the supporting structure was removed, at which time the
slab underwent a deflection of 0.2 to 0.3 in. These deflections
corresponded to the calculated values. The repairs were, therefore,
considered successful.
4. Kansas Department of Transportation. For many years, the Kansas
Department of Transportation (DOT) was faced with the problem of how to
repair shear cracks in the girders of many of its two-girder reinforced
concrete bridges. The Department had tried repairing some of the girders
using epoxy injection, but this method was not successful since the
repair procedure did not improve the shear capacity of the girder. Some
girders were repaired by removing the sections of cracked concrete in the
girder, adding reinforcing steel, and recasting the girder to its origi
nal dimensions. This method, however, requires the use of supporting
falsework and necessitates the closing of part of the bridge while the
work is being done.
64
As a result, an alternative method of repair was needed which would
permanently repair the cracks and improve the girders' shear capacity at
low cost and without serious traffic delays. Post-reinforcement, a re
pair method which meets each of these objectives, was subsequently devel
oped and is now routinely used by the Kansas DOT (Stratton and Crumpton,
1984; Stratton et al, 1978).
Post-reinforcement, as developed by the Kansas DOT, consists of the
following steps: (a) sealing of the surface of the crack using a sili
cone sealant, (b) vacuum drilling dust-free holes 6 in. apart and 45° to
the deck surface, thereby crossing the crack plane at 90°, (c) filling
the hole and crack plane with epoxy pumped under low pressure, and (d)
placing an almost full-depth length of reinforcing bar (No. 4 or 5) into
the drilled hole to span the crack by at least 18 in. The epoxy bonds
the bar to the walls of the hole and fills the sealed crack plane, which
bonds the cracked concrete surfaces together monolithically and rein
forces the section. A detailed description of the repair process, as
described by Stratton and Crumpton (19S4), is presented below.
Contractors are supplied construction plans and specifications for
performing a post reinforcement repair. If the bridge is covered
with an asphalt wearing surface, before drilling can begin, the
asphalt must be removed. Cracks on the surface of the girders are
sealed with an elastic silicone sealant. The clear sealant bead is
applied over the crack and then pressure screeded into the crack
with a specially shaped spreader. On wide cracks some buildup of
sealant may be needed. To avoid trapping rainwater in the cracks,
they are not sealed much in advance of the repair.
While sealing is in progress, the drill entry points are marked on
the bridge deck near the measured centerline of the girders. After
the ideal design positions are marked, the actual entry points are
established using a pachometer (metal detector) to find the trans
verse deck rebar. The actual drill entry position will usually be
within 2 in. of the design position shown on the plans. Without
the pachometer to locate the rebar, the drill crew stands a high
chance of hitting the rebar, which would cause tip breakage and
lost time.
65
Drilling should start at the center of the span and progress toward
the pier. This plan avoids having the drilling rig set up over holes
that are already drilled. To commence drilling, the trailer or truck
on which the drill is mounted is centered over the entry point, and
the drilling equipment is leveled side to side using hydraulic stabi
lizers. Next, the gantry that supports the drill is raised to a 45°
angle, locked at this angle, and then moved so the drill tip is at
the entry point marked on the bridge deck. A drill entry gauge can
be spot-faced in the deck using a light chipping hammer. This method
facilitates drill entry into the concrete and also reduces tip break
age. Once the tip is in position, the desired hole depth is marked
on the drill steel. Drilling the hole generally takes less than 1
minute, and turn-around time from hole to hole is usually less than
3 minutes. The number of holes drilled in a day is limited by the
number that can be injected with epoxy on the same day.
After being drilled, each hole is measured for depth, and a rebar is
cut 3 in. short of full hole depth. This length keeps the top of the
bar below the path of the grinder used to prepare decks topped with a
thin bonded concrete overlay.
The epoxy injection crew starts work after drilling is completed on
one girder part. Injection starts with the deepest hole, i.e., the
hole clo~est to the pier or the abutment. After the hole is about
half filled with epoxy, the bar is slowly inserted and gently tapped
to be sure it is seated on the bottom of the hole. The nozzle is
then reinserted, locked in place, and the hole is filled under
pressure.
Any crack in the girder will likely be intercepted by at least one
drilled hole. To fill tight cracks with epoxy, a sustained pump
pressure of 100 psi is usually required. Less pressure is required
to fill wide cracks. Consequently, the operator should not try to
increase pump pressure to 100 psi when wide cracks are encountered.
To do so might cause a rupture in the silicone material sealing the
crack exterior. Instead, the operator should watch the air cylinder
66
shaft of the pump for any displacement. As long as the air cylinder
shaft is moving and no leaks from the surface cracks are present,
the crack filling operation is progressing satisfactorily. Pumping
should continue until either epoxy is detected in the next hole or
the air cylinder shaft stops moving.
The injection crew should constantly check previously injected holes,
and if a lower level of epoxy is noted in the holes, they should be
refilled. In very hot weather, coarse aggregate can be poured into
the hole on top of the bar to cool the epoxy. Otherwise, the heat of
polymerization may cause the epoxy to boil and foam, or it may cause
thermal contraction cracks at the epoxy surface.
To perform the post reinforcement repair, a contractor must have cer
tain equipment. The l-in. diam holes must be clean, dust-free, and
dry. The depth of holes must be controlled, and the holes must be
straight to accept a No. 6 rebar. The drilling angle must be accu
rate and repeatable, and the rate of drilling should be fast. To
achieve these requirements, Kansas requires the use of a vacuum drill
that sucks up dust through the center of the hollow drill bit. A
proprietary trailer-mounted vacuum drill is available for less than
$40,000. A truck-mounted drill is available, too, but it costs more.
A pachometer, which usually costs less than $3,000, is needed to
locate the transverse bridge deck reinforcement. Without this
instrument, too many carbide drill tips would be broken and much
production time would be lost trying to locate holes by trial and
error.
The epoxy injection pump must also meet certain specifications.
The pump must be positive displacement and deliver a certified vol
ume ratio of hardener to resin in the temperature and pressure range
needed to perform the injection. It must be able to deliver a sus
tained pressure of 100 psi and must be controllable between 20 and
67
100 psi. The injection nozzle must lock in the hole and hold 100 psi
without leaking. The nozzle is a device which the contractor can build
himself or have a local machine shop build from a KsDOT design at minor
cost. Epoxy components must be kept separate and mixed just ahead of the
injection nozzle.
Stratton and Crumpton (1984) further report that an efficient crew
can completely post-reinforce a 3-span bridge in less than one week.
Since 1981, the Kansas DOT has used post-reinforcement to repair over
20 bridges at approximately $1,000 per girder, as compared to approxi
mately $40,000 per girder for girder removal and replacement.
It should be pointed out that while this procedure was specifically
developed for use with bridge deck beams and girders, there appears to be
no reason why the procedure cannot be modified for other applications.
68
PART IV: SU.HMARY AND RECOMMENDATIONS
According to the results of a survey initiated in 1982 by the U.S.
Army Corps of Engineers (McDonald and Campbell, 1985), the three most
common types of deficiencies encountered in concrete hydraulic structures
were (a) cracking, (b) seepage, and (c) spalling. These three general
catagoreis of deficiencies accounted for 77% of the 10,096 deficiencies
identified during a review of available inspection reports for the Corps'
civil works structures. Concrete cracking was the deficiency most often
observed, accounting for 38% of the total. I~ situ repair procedures may
not be readily applicable in the repair of seepage deficiencies; however,
the problems normally resulting from deterioration due to cracking and
spalling do seem to be suited to in situ repair procedures.
A literature survey, private discussions, and responses to mail and
telephone inquiries have disclosed a wide range of repair methods and ma
terials currently available for the in situ repair of cracked and spalled
concrete. Crack repair methods include pressure injection, routing and
sealing, stitching, addition of reinforcement, drilling and grouting,
flexible sealing, grouting, drypack mortar, crack arrest, polymer impreg
nation, and overlays and surface treatments. Methods for repairing
spalled concrete include coatings, concrete replacement, grinding, jack
eting, shotcreting, prepacked concrete, and thin-bonded or unbonded over
lays. Repair materials include bituminous materials, portland-cement
concrete, mortar and grouts, epoxies, expandable mortars, grouts and
concretes, linseed oil, latex-modified concrete, and polymer-concrete
materials.
From an evaluation of the repair techniques and materials identifi
ed, five procedures (three crack repair techniques and two techniques
for repairing spalled concrete) were identified as being the most appli
cable for in situ repair of concrete hydraulic structures. The selected
69
techniques include pressure injection, polymer impregnation, and addition
of reinforcement. In conjunction with these repair procedures, thin re
inforced overlays and shotcrete can be u~ed to repair spalled concrete
surfaces as well as to resurface a cracked structure after it has been
repaired.
Pressure injection, generally with epoxy adhesives, has been used
extensively for about 25 years to repair a variety of concrete hydraulic
structures. Consequently, much of the expertise, technology, materials,
and equipment necessary for successful application of the process already
exists. Advantages of pressure injection repair techniques include the
following: the cracks are sealed both internally and externally; by
proper selection of a water-compatible adhesive, cracks saturated with
water can be repaired; pressure injection can be used, within limits,
against a hydraulic head; and cracks as fine as 0.002 in. can be
repaired.
Pressure injection as a repair technique has the following limita
tions: the process is generally restricted to the repair of members that
have not yet begun to spall significantly and to the repair of dormant
cracks; and the process may leave scars on the surface of the member
where the cracks have been injected. Pressure injection, however, ap
pears to be one of the most viable methods available for repairing se
verely cracked concrete structures.
Limited use has been made of polymer impregnation over the last 8 to
10 years to rehabilitate highly deteriorated concrete structures. Most
of these applications have been experimental in nature, and routine ap
plications are not yet common, most likely because the process is rela
tively new and uses specialized materials and equipment requiring a high
level of expertise and supervision to ensure success.
Limitations of the polymer impregnation process include the follow
ing. The monomer systems currently in use require specialized safety
procedures since they are considered flammable and toxic. In addition,
they are not water compatible. Therefore, to obtain a complete cure of
70
the monomer system and to ensure adequate penetration of the monomer into
the pore structure of the concrete, it is necessary to dry the concrete
before impregnation in order to remove the free moisture within the
pores.
Despite these limitations, polymer impregnation appears to be one of
the best methods available for rehabilitating or improving the overall
physical and mechanical properties of highly deteriorated low-quality or
non-air-entrained concrete. Properties of polymer-impregnated concrete
which make it very attractive for use in concrete hydraulic structures
include low permeability to water and chloride penetration, high abrasion
resistance, excellent durability during cycles of freezing and thawing,
and compressive and flexural strengths three to four times greater than
those of ordinary concrete.
The addition of reinforcement to deteriorated concrete structures,
either internally or externally, has also been shown to be an excellent
means for repairing cracked structures, particularly when it is also
desirable to improve or restore the strength properties of the members
being repaired. Both methods of repair have been used extensively for a
number of years. As a result, much of the necessary materials, equipment
and expertise necessary for the successful application of the repair
techniques already exist.
Once the cracks in a member have been repaired, it may be necessary
to resurface the member to (a) cover up any scars or imperfections left
by the repair procedure, (b) repair minor damage resulting from spalling,
or (c) provide the member with a more durable wearing surface. Thin re
inforced overlays, either bonded or unhanded, and shotcrete appear to be
the best methods available. Conventional portland-cement mortar systems
or concrete-polymer material systerns can be used for either resurfacing
method.
Although each of the selected repair techniques has definite advan
tages over other forms of repair, it is felt that they could be improved
to make them more applicable to the repair of hydraulic structures.
71
Areas which should be investigated include the following:
(a) Development of a repair technique which uses both pressure in
jection and polymer impregnation. Many of the older structures which are
cracked contain concrete that is of low qualiLy or is non-air entrained.
While pressure injection repair techniques will effectively seal the
crack network and rebond the concrete, the concrete is still susceptible
to further cracking because the overall quality of the concrete itself
remains unchanged. Impregnation of the concrete would improve the physi
cal and mechanical properties of concrete, thereby reducing the possibil
ity of further deterioration.
(b) Identification of water compatible monomers for use in impreg
nation work. At present, the monomers being used to impregnate concrete
are not compatible with water. It is, therefore, necessary to dry the
concrete to remove the free moisture from its pore system to ensure
adequate penetration of the monomer. If this step could be eliminated,
the time and cost of the repair process could be reduced. These monomers
could also be used in pressure injection repair techniques provided a
suitable initiator-promoter system can also be identified.
(c) Identification of low vapor pressure monomers for use in im
pregnation work. These monomers would help to eliminate some of the
safety problems such as toxicity and flammability associated with the
higher vapor pressure monomers currently being used.
(d) Develop and refine field impregnation techniques. Since most
of the impregnation work performed to date has primarily been on hori
zontal surfaces, it will be necessary to develop impregnation techniques
for use on curved and vertical surfaces such as those on concrete piers.
(e) Investigate the use of spray applicable polymer concrete over
lays. Sprayable polymer concrete overlays can be used to resurface a
concrete member once it has been repaired or to seal the concrete, there
by reducing its permeability. Fillers, such as calcined coke breeze, can
be added to the overlay to make it electrically conductive. The overlay
can then be used as the anode in an impressed current cathodic protection
system to prevent reinforcing steel from corroding.
72
(f) Investigate the use of nonmetallic reinforcement in thin over
lays. One problem with the use of thin overlays is that they are subject
to reflective cracking. The use of nonmetallic reinforcing, such as
polymer grids, may help to eliminate this problem.
(g) Work may be needed to adapt the Kansas Department of
Transportation's post-reinforcement repair techniques to nonhorizontal
applications.
73
REFERENCES
ACI Committee 224, 1984. "Causes, Evaluation, and Repair of Cracks in Concrete Structures," Repcrt No. 2241.R-84, ACI Journal, May-June, 211-230.
Adhesives Engineering Bulletin, 1971. "Leaking Cracks In Twi1: Lakes Dam are Sealed with Injected Epoxy," VI(2), May, !+,
Adhesives Engineering Bulletin, 1973. "Cracks in Spillway Tunnel of Dam Sealed Fully with Injected Epoxy," VIII(2), May, 1.
Adhesives Engineering Bulletin, 1981a. "Major Cracks in Canadian Dam Repaired By Adhesive Injection," 16(1), Nov., 1-2.
Adhesives Engineering Bulletin, 1981b. "Steel Tendons, SCB Injection Stop Leaks In Navigation Lock," 16(1), Nov., 6.
Barlow, P., 1986. "Repairing a Major Concrete Navigation Lock," Concrete International, 8(4), Apr. 1986, 50-52.
Concrete Construction, 1982. "Concrete Repair: Methods and Materials," 54 p.
Flock, R. and Walleser, M., 1983. sippi River, Canton, MO., Dam No. Injection Pier No. 39," u.s. Army District.
"Nine-Foot Channel Project, Missis-20, Concrete Rehabilitation, Epoxy Corps of Engineers, Rock Island
Fontana, J.J. and Kukacka, L.E., 1979. "Greenport Bridge Reconstruction Using Concrete Polymer Materials," BNL-26353, Brookhaven National Laboratory, Upton, NY.
Kaeding, A.O., 1976. "Building Restoration Using Monomer Impregnation," Presented at the 1976 Fall Convention of the American Concrete Institute, Mexico City, Mexico, Oct. 24-29, 1976.
Kaeding, A.O., 1978. "Structural Use of Polymers in Concrete," In: Proceedings of the Second International Congress of Polymers in Concrete, Univ. of Texas, Austin, TX, Oct. 25-27, 1978, 9-23.
Keith, J. M., 1972. "Repair of Culvert Valve Monoliths Using Rock Anchors," Presented by ORLED at the Structural Conference in St. Paul, MN, July, 1972.
McDonald, J. E. and Campbell, R. L., 1985. "The Condition of Corps of Engineers Civil Works Concrete Structures," Technical Report REMR-CS-2, U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS.
74
REFERENCES - Cont.
Murray, M.A. and Schultheis, V.F., 1977. "Polymerizalion of Concrete Fights Cavitation," Civil Engineering - ASCE, April, 67-70.
Neuberger, K. J., 1982. "John Day Lock Repair," Concrete Structures, Repair, and Rehabilitation, Vol. C-82-1, Sept. 1982, 1-4.
Scanlon, J.M. et al., 1983. "REMR Research Program Development Report, Final Report," U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS.
Schrader, E.K. and Kaden, R.A., 1976. "Stilling Basin Repairs at Dworshak Dam," The Military Engineer, NO. 444, 282-286.
Schrader, E.K., 1978. "The Use of Polymer in Concrete to Resist Cavitation/Erosion Damage," In: Proceedings of the Second International Congress on Polymers in Concrete, Univ. of Texas, Austin, TX, Oct. 25-27, 1978, 283-309.
Stratton, F.W., Alexander, R., and Nolting, W., 1978. "Kansas Bridges Renovated by Post Reinforcement and Thin Bonded Concrete Overlay," Concrete Construction, Aug., 705 and 707-709.
Stratton, F. W. and Crumpton, C. F., 1984. "Kansas Bridges Renovated by Post Reinforcement and Thin Bonded Concrete Overlay," Concrete Construction, Aug., 705 and 707-709.
Van Gernert, D. and Maesschalck, R., 1983. "Structural Repair of a Reinforced Concrete Plate by Epoxy Bonded External Reinforcement," Int. J. Chern. Compos. Lightweight Concr., 5(4), 247-255.
75
Table 1
Crack Repair Techniques for Concrete
Repair Technique
Pressure Injection
Routing and Sealing
Stitching
Addition of Reinforcement
Drilling and Grouting
Type of Crack Dormant Active
X
X
X X
X X
X
76
Comments
Little surface preparation is needed; scar marks may be left on surface where crack was injected. Limited to areas where concrete has not yet spalled. Structural quality bond is established but if large structural movements are still occurring, new cracks may open. Process can be used against a hydraulic head.
Simplest method available for repair of cracks with no structural significance. Process not applicable to repair of cracks subjected to hydraulic head.
Process will not close or seal cracks but can be used to prevent them from progressing. Generally used when it is necessary to reestablish tensile strength across crack.
Primarily used to restore or upgrade structural properties of cracked members.
Technique applicable only when cracks run in straight line and are accessbile at one end.
Repair Technique
Flexible Sealing
Grouting
Drypack Mortar
Crack Arrest
Impregnation
Overlays and Surface
Treatments
Autogenous Healing
Table 1 cont.
Crack Repair Techniques for Concrete
Type of Crack Dormant Active
X X
X
X
X X
X X
X X
X
77
Comments
Technique is applicable where
appearance is not important
and in areas where cracks are
not subjected to traffic or
mechanical abuse.
Wide cracks may be filled
with portland-cement grout.
Narrow cracks may be filled
with chemical grouts.
For use in cavities that are
deeper than they are wide.
Convenient for repair of
vertical members.
Commonly used to prevent prop
agation of cracks into new
concrete during construction.
Technique can be used to re
store structural integrity of
highly deteriorated or low
quality concrete. Can be
used to seal small crack networks.
Slabs containing fine dormant
cracks can be repaired using
bonded overlays. Unbonded
overlays should be used to
cover active cracks.
A natural process of crack
repair has practical applica
tions for closing dormant
cracks in moist environments.
Table 2
Techniques for Repairing Spalled Concrete
Repair Tec~nique
Coatings
Concrete Replacement
Grinding
Jacketing
Shotcreting
Prepacked Concrete
Thin-Bonded and
Unbonded Overlays
Comments
This technique is generally used when
the scaling or spalling is limited to a
very thin region at the surface of the
concrete.
This technique is one of the most com
monly used and is appropriate for appli
cations where the cause of deterioration
is nonrepeating or has been eliminated.
This technique can be used when the de
terioration is limited to a thin region
at the surface of the concrete.
This technique entails fastening a ma
terial to the existing concrete that is
more resistant to the environment that
is causing the deterioration.
This technique is practical for large
jobs, on either vertical or horizontal
surfaces, where the cavities are rela
tively shallow.
This technique is suitable for inacces
sible applications, such as submerged
concrete or deteriorated concrete that
is being jacketed.
Thin overlays are often used to repair
surfaces that are basically sound struc
turally but have deteriorated because
of cycles of freezing and thawing, heavy
traffic, or other exposures which the
original concrete was unable to
withstand.
78
Table 3
Materials for Repairing Spalled Concrete
Repair Material
Bituminous Coatings
Concrete, Mortar, or Grout
Epoxies
Expanding Mortars, Grouts,
and Concretes
Linseed Oil
Latex-Modified Concrete
Comments
Asphalt- or coal-tar-based bituminous
coatings are used to waterproof concrete
or protect it, to some extent, from
weathering.
Portland-cement concrete, mortar, and
grout have a number of advantages as a repair material, including: thermal prop
erties similar to the existing concrete,
similarity in appearance, comparatively
low cost, availability, and familiarity.
Epoxies are most often employed in repair
work for the following uses: as an adhe
sive to bond plastic concrete to hardened
concrete or other rigid materials, for patching, and for coating concrete to
protect it from aggressive environments.
These materials are generally proprietary materials to counteract the problem of
shrinkage by incorporating ingredients which produce an expansive force approxi
mately equal in magnitude to the shrink
age stresses.
Linseed oil is generally used to prevent
or minimize additional scaling from
occurring.
Latex-modified concretes have generally
been used for resurfacing deteriorated
floors and bridge decks. They typically
develop higher strengths, bond better to
existing concrete, have higher resist
ances to chloride penetration, and are
more resistant to chemical attack than
plain concrete.
79
Table 3 cant.
Materials for Repairing Spalled Concrete
Repair Material
Polymer Concrete
Comments
Polymer concrete has been used exten
sively to repair highway bridges and
pavements. It has a number of advantages
over normal concrete, including rapid cur
ing characteristics, high early strength,
good bond strength, and excellent durabil
ity through cycles of freezing and
thawing.
80